Improvement of Cell Permeability of Human Neuronal Nitric Oxide Synthase Inhibitors Using Potent and Selective 2-Aminopyridine-Based Scaffolds with a Fluorobenzene Linker

Article

pubs.acs.org/jmc

Cite This: J. Med. Chem. XXXX, XXX, XXX-XXX

Improvement of Cell Permeability of Human Neuronal Nitric Oxide

Synthase Inhibitors Using Potent and Selective 2‑Aminopyridine-

Based Scaﬀolds with a Fluorobenzene Linker

Ha T. Do,^†,§Heng-Yen Wang,^†,§Huiying Li,^‡Georges Chreiﬁ,^‡Thomas L. Poulos,*^,‡

and Richard B. Silverman*^,†

^†Department of Chemistry, Department of Molecular Biosciences, Chemistry of Life Processes Institute, Center for Molecular

Innovation and Drug Discovery, Center for Developmental Therapeutics, Northwestern University, 2145 Sheridan Road, Evanston,

Illinois 60208-3113, United States

^‡Departments of Molecular Biology and Biochemistry, Pharmaceutical Sciences, and Chemistry, University of California, Irvine, Irvine

California 92697-3900, United States

S

* Supporting Information

ABSTRACT: Inhibition of neuronal nitric oxide synthase (nNOS) is a promising therapeutic approach to treat

neurodegenerative diseases. Recently, we have achieved considerable progress in improving the potency and isoform selectivity

of human nNOS inhibitors bearing a 2-aminopyridine scaﬀold. However, these inhibitors still suﬀered from too low cell

membrane permeability to enter into CNS drug development. We report herein our studies to improve permeability of nNOS

inhibitors as measured by both PAMPA−BBB and Caco-2 assays. The most permeable compound (12) in this study still

preserves excellent potency with human nNOS (K_i= 30 nM) and very high selectivity over other NOS isoforms, especially

human eNOS (hnNOS/heNOS = 2799, the highest hnNOS/heNOS ratio we have obtained to date). X-ray crystallographic

analysis reveals that 12 adopts a similar binding mode in both rat and human nNOS, in which the 2-aminopyridine and the

ﬂuorobenzene linker form crucial hydrogen bonds with glutamate and tyrosine residues, respectively.

Although NO possesses many essential functions in cells, its

overproduction has been implicated in various pathological

disorders.⁶⁻⁹In the central nervous system (CNS), excess NO

produced by nNOS has been linked to diverse neuronal

disorders such as Parkinson’s disease, Alzheimer’s disease,

Huntington’s disease, amyotrophic lateral sclerosis, ischemic

stroke, and migraines.^6,10,11At high concentrations, NO causes

excessive nitration and/or nitrosylation of proteins, leading to

their degradation and misfolding.¹²Moreover, overproduced

NO can also react with superoxide anion to form a highly

reactive oxidant, peroxynitrite, which causes DNA damage, lipid

peroxidation, and mitochondria dysfunction and therefore

triggers neuronal apoptosis or necrosis.¹³These processes

lead to synaptic damage and neurotransmission impairment,

which are commonly observed in symptoms of neuro-

degenerative diseases.¹⁴Consequently, limiting NO over-

INTRODUCTION

■

Nitric oxide (NO) is a unique cell signaling molecule involving

in many physiological functions including neurotransmission,

vasodilation, smooth muscle relaxation, vascular regulation, and

immune response.¹⁻³NO is generated in cells by a family of

enzymes called nitric oxide synthase (NOS), which catalyzes

the oxidation of ^L-arginine (L-Arg) to L-citrulline. Three

diﬀerent isoforms of NOS have been found in mammals, of

which two isoforms, neuronal NOS (nNOS) and endothelial

NOS (eNOS), are constitutively expressed in relation to the

calcium levels in cells, while the third one, inducible NOS

(iNOS), is expressed in response to cell inﬂammation.⁴These

isoforms are distributed in various tissues and responsible for

diﬀerent physiological functions: nNOS is mainly localized in

the nervous system and is involved in neuronal communication,

eNOS is found in the endothelium and participates in

regulation of vascular pressure and vasodilation, and iNOS is

located in macrophages and is essential for the innate immune

system.⁵

Received: September 12, 2017

A

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 1. Reported nNOS inhibitors with some improved pharmacokinetic properties.

Figure 2. Structural modiﬁcation of lead compound 6 to improve its permeability

production through the inhibition of nNOS is a potential

therapeutic approach for the treatment of neuronal disorders.

NOS isoforms exist as homodimers, in which each monomer

consists of one N-terminal oxygenase domain and one C-

terminal reductase domain connected to each other by a

calmodulin binding region. The former domain contains a

noncatalytic zinc, tetra-hydrobiopterin (H₄B), and a heme-

containing catalytic active site that binds the substrate, ^L-Arg.

The latter domain, the C-terminal reductase, employs

nicotinamide adenine dinucleotide phosphate (NADPH), ﬂavin

adenine dinucleotide (FAD), and ﬂavin mononucleotide

(FMN) to shuttle electrons to the iron center of the heme

cofactor in the catalytic active site, where ^L-Arg gets oxidized to

L-citrulline and releases NO.⁴Because of the high similarity in

the structure of the active sites of the three NOS isoforms and

the diverse involvement of NO signaling in many cellular

physiological functions, selective inhibition of nNOS over

eNOS and iNOS to develop a new therapeutic with minimal

side eﬀects for neurodegenerative diseases is a critical

challenge.¹⁵Moreover, the presence of the blood−brain barrier

(BBB) formed by endothelial cells with tight junctions creates

an additional challenge for CNS drug development, as it greatly

limits the delivery of drugs into the brain.¹⁶

as incorporating intramolecular hydrogen bonds (1),¹⁸

converting to prodrugs (2),¹⁹and modulating the amine

basicity (3),²⁰and replacing an essential pharmacophore, such

as replacement of the 2-aminopyridine with a 2-imidazolylpyr-

imidine (4)²¹or 2-aminoquinoline (5)²²(Figure 1). Although

some pharmacokinetic improvement has been achieved, these

approaches can result in a diminution in activity, especially for

human nNOS, the ultimate target for neurodegenerative

diseases, and in the selectivity over human eNOS.

Recently, we reported a series of nNOS inhibitors containing

a 2-aminopyridine anchor attached to a pyridine ring linker and

an amine tail (6, Figure 1).²³Compared to other nNOS

inhibitors, 6 is considerably easier to synthesize and possesses

excellent potency toward rat and human nNOS (K_i(rnNOS) =

16 nM; K_i(hnNOS) = 13 nM) as well as high selectivity over

iNOS and heNOS (n/i = 116 and hn/he = 1831). Additionally,

pharmacokinetic studies revealed that this compound displayed

very poor CYPs inhibition, little human microsome metabo-

lism, and only 20% human plasma protein binding. Never-

theless, 6 expresses little Caco-2 permeability, an indication of

poor predicted permeation through the BBB, thereby rendering

low therapeutic value.

Our further investigations, reported herein, have been

focused on improving the permeability of 6 while retaining

its excellent potency and selectivity. In previous reports,^23,24we

demonstrated that the 2-aminopyridine moiety was crucial for

the key interactions of nNOS inhibitors with Glu-592 and Glu-

597 in the active site of rnNOS and hnNOS, respectively.

Hence, structural modiﬁcations have been focused on the

middle ring linker and the amine tail. First, the middle pyridine

linker was replaced by other aromatic rings. Various linker

moieties (7−10) (Figure 2) were chosen to explore the activity

Nevertheless, a large number of compounds with good

potency and high selectivity for inhibiting nNOS over eNOS

and iNOS have been reported. Most of them are designed to

compete with ^L-Arg binding at the active site of the enzyme.¹⁷

These inhibitors, however, still suﬀer from low membrane

permeability and poor ability to cross the BBB as a result of

their high basicity and/or polarity. In the past few years our

group has pursued various medicinal chemistry approaches to

improve the pharmacokinetic values of nNOS inhibitors, such

B

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 1

a

Reagents and conditions: benzylic bromide 21a, 21b, 21c, or 21d, n-BuLi, THF, −78 to 0 °C, then −78 °C.

a

Scheme 2

a

Reagents and conditions: (a) 23a or 23b, Pd(PPh₃)₄, CuI, TEA, 90 °C, 20 h; (b) 20% TFA in CH₂Cl₂, rt, 1 h; (c) (i) Pd/C, H₂, MeOH, rt, 20 h,

(ii) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h.

a

Scheme 3

a

Reagents and conditions: (a) CuCN (1 equiv), pyridine (1 equiv), DMF, 150 °C; (b) 23a or 23b, Pd(PPh₃)₄, CuI, TEA, 90 °C, 20 h; (c) 20% TFA

in CH₂Cl₂, rt, 1 h; (d) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h; (e) Pd/C, H₂, MeOH, rt, 20 h.

as well as enhance the lipophilicity. Second, methylation of the

secondary amine (11−14) (Figure 2) was carried out in a

strategy to improve the permeability by reduction in the

number of H-bond donors. Third, reduction of the number of

rotatable bonds within the molecule was investigated by

incorporating biaryl linkers bearing an amino tail at both

meta- (16) and para- (17) positions as well as modulation of

the basicity of the amino tail group in molecules containing

these rigid linkers (15). Finally, introduction of lipophilic

moieties to the biaryl linkers (18,19) (Figure 2) was employed

to maximize the lipophilicity of the modiﬁed compounds. All

compounds were then investigated for their nNOS inhibition

and selectivity over eNOS and iNOS. Analogues with high

potency and selectivity were further examined in a parallel

artiﬁcial membrane permeability for blood-brain barrier

(PAMPA−BBB) assay for their permeability and in a Caco-2

assay for their P-glycoprotein (P-gp) substrate liability.

RESULTS AND DISCUSSION

■

Chemistry. To replace the pyridine linker in 6 by other

aromatic rings, diﬀerent benzylic bromide compounds (21a−

d), which are commercially available, were substituted with

lithiated pyrrolyl-4,6-dimethylpyridine (20) to generate inter-

mediates 22a−d (Scheme 1).

The generated intermediates (22a−c) were then coupled

with either N-Boc-N-methyl-propargylamine (23a) or 3-

dimethylamino-1-propyne (23b) via Sonagashira cross-cou-

C

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 4

a

Reagents and conditions: (a) boronic acid (31a, 31b, or 31c), K₂CO₃, Pd(PPh₃)₄, THF/water (4/1), 80 °C, 20 h; (b) 20% TFA in CH₂Cl₂, rt, 1 h;

(c) (i) N-methylamine hydrochloride salt, NaOAc, MeOH, rt, 15 min, (ii) NaBH₄, MeOH, 0 °C to rt, 1 h; (d) NH₂OH·HCl, EtOH/H₂O (2:1), 100

°C, 20 h.

pling to aﬀord alkynes carrying Boc-protected secondary amine

(24a−c) or tertiary amine (26a−c), respectively. A sequence of

Boc deprotection, alkyne reduction, and pyrrole deprotection of

24a−c provided desired compounds 7−9. In a similar pathway

of alkyne reduction and pyrrole deprotection, compounds 11−

13 were obtained from 26a−c (Scheme 2).

To maximize the possible increase in cell membrane

permeability, a new set of compounds that contains both

lipophilicity and rigidity enhancement were synthesized. These

compounds were designed to consist of a biaryl linker

containing a ﬂuorophenyl ring. Following a similar synthetic

route used for analogues 15−17, compounds 18 and 19 were

successfully constructed using 22b in place of intermediate 30

(Scheme 4).

It was desirable to synthesize compounds 10 and 14,

containing a cyanophenyl linker, since our previous studies

showed that incorporation of a cyano group into potential

molecules helps improve their nNOS activity and selectivity,

especially with human nNOS.^22,24Intermediate 27, containing

a cyanophenyl linker, was synthesized from bromophenyl

precursor 22d by treatment with CuCN in DMF at 150 °C.

Sonogashira coupling was then performed on 27 to install the

amine tails. Unlike the synthetic route for 7−9 and 11−13,

pyrrole deprotection in the synthesis of target compounds 10

and 14 was performed before alkyne reduction to avoid

overreduction of the pyrrole ring by Pd/C, H₂(Scheme 3).

The syntheses of compounds containing pyridine-based

biaryl linkers were started with construction of the biaryl

moiety using Suzuki coupling of 30 with diﬀerent boronic acids

(31a−c) as shown in Scheme 4. Two assessments were

investigated in this modiﬁcation with pyridine-based biaryl

linkers. First, the boronic acid of Boc-protected aniline 31a was

used to modulate the basicity of the tail amino group.

Reduction of the pK_anot only improves the permeability of

the resulting compound but also reduces its P-gp substrate

liability.²⁵Second, compounds in which the amino group is

connected to the pyridine-based biaryl linkers via one extended

methylene group at both meta- or para- positions were

synthesized to evaluate the eﬀect of these connectivities on

the biological activity and selectivity of the resulting inhibitors.

Boc-deprotection of 32a obtained from the aforementioned

Suzuki reactions yielded intermediate 33a, while reductive

amination of 32b,c with the N-methylamine hydrochloride salt

generated 33b,c. Pyrrole deprotection of the three intermedi-

ates (33a−c) aﬀorded the ﬁnal desired products (15−17,

Scheme 4).

Biological Activity. The hemoglobin NO capture assay²⁶

was used to determine the inhibitory constants (K_i) of

synthesized compounds 7−19. First, these compounds were

tested against rat nNOS and murine iNOS to evaluate their

inhibitory activity and isoform selectivity. On the basis of the

screening results, the compounds with high potency and

selectivity were further assayed against human nNOS and

human eNOS. The K_iand the selectivity values of 7−19, as well

as those of compound 6 for comparison, are summarized in

Table 1. The selectivity for nNOS over iNOS (n/i) was

obtained by comparing K_ivalues of rat nNOS and murine

iNOS on account of the ease of expression and puriﬁcation of

these enzymes. For selectivity against nNOS over eNOS, K_i

values obtained from human nNOS and human eNOS were

used (hn/he) to achieve structure−activity relationships (SAR)

closer to the human system. Additionally, a cross comparison of

compound activities against nNOS of diﬀerent species (human

vs rat, hn/rn) was also evaluated because this is valuable

information in evaluating translation from preclinical to clinical

studies.

In the ﬁrst series of compounds, modiﬁed by replacing the

middle pyridine linker with other aromatic rings (7−10),

modiﬁcation caused eﬀects on both potency and selectivity.

Particularly, when phenyl or ﬂuorophenyl linkers were

employed in place of pyridine, the potencies with rat nNOS

and human nNOS of the resulting compounds (7 and 8,

respectively) are slightly decreased compared to those of 6.

When the aromatic linkers contain a bulkier and more electron

withdrawing group, i.e., triﬂuoromethyl- (9) or cyano- (10)

groups, the resulting compounds display a dramatic drop in the

D

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Table 1. List of K_iValues and Selectivity of 6−19

K_i(nM)

selectivity

rat

human

nNOS

murine

iNOS

human

eNOS

compd nNOS

n/i

hn/he hn/rn

6

16

48

13

55

1894

3614

4649

33690

16165

3458

3857

36017

16373

6704

3005

4213

3012

10883

22895

20117

39505

51542

42086

20816

83976

NT

118

75

113

98

64

58

149

67

71

4

1761

365

610

86.1

81.5

485

2799

ND

106

685

234

193

0.8

1.1

1.6

1.7

2.0

0.7

1.2

ND

1.7

1.8

1.9

2

7

8

41

65

9

346

252

60

598

516

43

10

11

12

13

14

15

16

17

18

19

26

30

541

232

1700

47

NT

82

NT

8668

64

74

69

59

57

100

83

68539

19415

69888

44

185

362

a

K_ivalues were calculated from the IC₅₀values of a dose−response

curve using the Cheng−Prusoﬀ equation.²⁷6−9 concentrations were

tested and the IC₅₀is an average of at least two duplicates. The

standard error is less than 10%. NT: not tested. ND: not determined.

inhibitory activity over nNOS in both species. This result

suggests a diﬀerent eﬀect from the modiﬁcation of the middle

aromatic ring compared to what we observed previously for the

other pharmacophores such as a 2-aminoquinoline anchoring

head with an amine tail or a 2-aminopyridine with a diamine

tail;^22,24in both cases, incorporation of a nitrile on the middle

phenyl ring generally resulted in improvement in both potency

and selectivity. This trend, however, does not consistently

occur in the current scaﬀold of nNOS inhibitors, which consists

of a 2-aminopyridine anchoring head and a middle aromatic

ring with a long amine tail.

X-ray crystal structures of these inhibitors bound to NOS

were determined to reveal the structural basis for these new

observations. It has been shown that²³the middle pyridine ring

of 6 can establish an upward binding mode with the nitrogen

atom forming a H-bond to a Tyr residue (Tyr-562 in rnNOS

and Tyr-567 in hnNOS). Once the middle pyridine is replaced

by a phenyl ring in 7, this H-bonding capability is removed. As

shown in the structure of rnNOS-7 (Supporting Information,

Figure S1A), while anchoring the 2-aminopyridine can still

make bifurcated H-bonds with Glu-592, the middle phenyl ring

bends back on to the heme, and the long amine tail does not

make any favorable contacts with the protein. A similar

situation is also observed in the hnNOS-7 structure

(Supporting Information, Figure S1B). As the result of these

changes in the interactions with nNOS, the nNOS inhibitory

potency of 7 drops by 3−4-fold compared to those of 6 (Table

1).

The upward binding mode of the middle aromatic ring is

reestablished when the phenyl ring is replaced by a

ﬂuorophenyl ring. The crystal structures of 8 bound to both

rnNOS and hnNOS (Figure 3A,B, respectively) clearly show

the H-bond from the ﬂuorine atom to Tyr-562 (rnNOS, 2.7 Å)

or Tyr-567 (hnNOS, 2.5 Å). In addition, the tail amine moiety

can also H-bond (2.6−2.8 Å) with the water molecule bridging

between the heme propionate A and H₄B. The “3-points” H-

bonding interactions (i.e., head, middle, and tail) shown with

compound 8 are similar to what were seen for 6, although the

binding aﬃnity of 8 to both nNOS enzymes is 3−5-fold weaker

Figure 3. Active site structure of 8 bound to rnNOS (A), hnNOS (B),

and heNOS (C). For this and all following structural ﬁgures, major H-

bonds are depicted with dashed lines and distances are labeled in Å.

The F_o− F_comit electron density for the bound inhibitor is contoured

at 2.5 σ. All of the structural ﬁgures were prepared with PyMol (www.

pymol.org).

than that of 6, possibly the result of the nonpolar nature of the

ﬂuorophenyl ring in 8 versus the polar pyridine ring in 6.

Compound 8 also retains good isoform selectivity compared

to lead compound 6 (6, n/i = 118, hn/he = 1761; 8, n/i = 113,

hn/he = 610). The structure of heNOS-8 (Figure 3C) indicates

that while the 2-aminopyridine group can still H-bond with

Glu-361, the middle ﬂuorophenyl ring and the tail amine show

a high degree of uncertainty with incomplete electron densities.

The ethylene linker between the 2-aminopyridine and the

middle phenyl ring is not pointing upward, as seen in the

structures of 8 bound to nNOS (Figure 3A,B). As a result, the

ﬂuorophenyl ring no longer H-bonds with Tyr331. Moreover,

the disordered tail amine does not make any H-bond with

protein or water. Altogether, the lack of these interactions could

be responsible for the low inhibition of this compound against

heNOS, which results in its high hn/he ratio.

The ratio of activity between hnNOS and rnNOS of 7−10

(hn/rn) was also determined to evaluate the potential

translation of these inhibitors from preclinical data to a clinical

E

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

study. This ratio was aimed to be as close to 1.0 as possible so

that there will be little to no signiﬁcant diﬀerence in the amount

of inhibitors used in rat and human dosage. The primary

sequence of human nNOS and rat nNOS is more than 93%

identical, and the nNOS active site is quite conserved between

these two species.²⁴The diﬀerence between them that is

relevant to inhibitor binding in a peripheral pocket is that the

nonpolar Leu337 residue in rnNOS is replaced by a bulkier and

polar His342 residue in hnNOS. Compounds 7−10, which

contain a nonbulky alkylamino tail were designed not to reach

into this pocket, therefore resulting in a close-to-1.0 hn/rn

ratio. As shown in Table 1, although 7−10 displayed slightly

lower inhibition to human nNOS than rat nNOS, their hn/rn

ratio is still less than 2.0, especially, 7 and 8, the two favorably

potent and selective compounds.

Our second approach to improve the permeability of 6, in

which the secondary amino tail of 7−10 was replaced by a

tertiary amino tail to reduce the number of hydrogen bond

donors in the lead molecule, gave the next series of compounds

(11−14). In the previous study,²³the H-bonds of the

secondary amino tail group of 6 with a bridging water molecule

interacting with both the heme and H₄B moieties of nNOS was

proposed to account for the improvement in potency of 6 with

hnNOS. Hence, the conversion of the secondary amine in the

tail of 7−10 to a tertiary amine in the corresponding analogues

(11−14) may potentially diminish the hnNOS inhibitory

activity of these resulting compounds. However, we found that

the secondary-to-tertiary amine conversion had little eﬀect on

the inhibition activity of 11−14 compared to those of 7−10. In

fact, 12 is more potent with both rat and human nNOS than its

secondary amine analogue (8), with K_ivalues close to those of

lead compound 6. Moreover, 12 displays a remarkable

improvement in isoform selectivity surpassing those of 6,

especially over heNOS (hn/he = 2799). Additionally, 12 also

has a favorable hn/rn ratio (1.2). X-ray crystal structures of 12

bound to rnNOS, hnNOS, and heNOS are shown in Figure 4.

Similar to lead compound 6 and its secondary amine

analogue 7, compound 12 retains the upward binding mode in

both rnNOS (Figure 4A) and hnNOS (Figure 4B), with its

ﬂuorophenyl ring forming a H-bond with a Tyr residue

(Tyr562 in rnNOS or Tyr567 in hnNOS). The tertiary amine

can still approach the bridging water molecule between heme

propionate A and H₄B. The only diﬀerence from 7 is that the

tertiary amino group of 12 is now a H-bond acceptor instead of

the donor of 7. This bridging water is not visible in the hnNOS-

12 structure because of the moderate data resolution (2.45 Å)

but is supposed to be there, interacting with the tertiary amino

group of 12.

Compared to 8 and 6, 12 exhibited a much greater

improvement in the selectivity for hnNOS over heNOS (4.5

times more than 8 and 1.6 times more than 6). The X-ray

structure of 12 bound to heNOS (Figure 4C) reveals that the

ﬂuorophenyl linker adopts a bent-over binding mode with the

ring pressing against the heme moiety, which is diﬀerent from

the structure of the heNOS-8 complex, in which the middle

ring is ﬂipped over 180° with the ﬂuorine atom pointing toward

Asn366. The tertiary amino group is not making H-bond or

other favorable contacts with the protein, resulting in a 2-fold

weaker potency than 8. Therefore, compound 12 becomes the

most selective inhibitor in the series with hn/he = 2799.

As the K_ivalues listed in Table 1 indicate that a

triﬂuoromethylphenyl (9 and 13) or a benzonitrile (10 and

14) as the middle ring leads to poor inhibitors, we attempted to

Figure 4. Active site structures of 12 bound to rnNOS (A), hnNOS

(B), and heNOS (C). The absence of the H₄B site water in the

hnNOS structure is likely the result of the moderate data resolution.

look for the underlying reasons in crystal structures. We did not

obtain decent data for rnNOS-9, but data for both rnNOS-13

(Supporting Information, Figure S2A) and hnNOS-13

(Supporting Information, Figure S2B) allowed us to model

the inhibitor in the active site. The 2-aminopyridine group H-

bonds with Glu-592 (rnNOS) or Glu-597 (hnNOS), and the

ethylene linker connected to the middle aromatic ring also

bends upward as that observed in the rnNOS-12 structure.

However, because of the bulky triﬂuoromethyl group, the

phenyl ring itself cannot go upward; instead, the triﬂuor-

omethyl group ﬁts into a pocket between Glu-592 and Arg-596

in rnNOS, making nonbonded interactions with the two side

chains. The long amine tail is highly ﬂexible with weak density,

approaching heme propionate D but without forming a H-

bond. Losing H-bonds from the middle ring and the tail amine

in 13 (or 9) are likely the reasons behind their poor binding.

Using a benzonitrile ring as the middle aromatic linker had

some success in the 2-aminoquinoline-²²and the 2-amino-

pyridine-²⁴containing inhibitors in the past, but it showed the

opposite eﬀects with compounds 10 or 14 here. We were able

to obtain the structures of rnNOS-10 (Figure S3A) and

hnNOS-10 (Figure S3B). In these structures, the benzonitrile

ring position in 10 is similar to that found for the ﬂuorophenyl

ring in 8, with the cyano group protruding next to Asp-597

F

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 5. Active site structures of 16 bound to rnNOS (A) and hnNOS (B).

Figure 6. Active site structures of 17 bound to rnNOS (A) and hnNOS (B).

(rnNOS) or Asp-602 (hnNOS) at a distance of 2.9−3.0 Å. This

interaction cannot be a H-bond unless the Asp residue is

protonated. The tail amino group is ﬂexible but managed to

make a H-bond with the bridging water near the H₄B. Given

the poor inhibitory potency, the close contacts involving the

by ChemAxon software, https://www.chemaxon.com/). The

assay with rnNOS showed that the reduction in the basicity of

the tail amino group drastically reduced the nNOS inhibition of

15, which is 100 times less potent than 6. No crystal structures

were determined with nNOS for compound 15 because of its

poor binding aﬃnity.

cyano group may even be considered as unfavorable clashes.

Why would a middle benzonitrile ring make 2-aminoquino-

line compounds better inhibitors²²but not the 2-aminopyridine

analogues described here? The bulkier 2-aminoquinoline group

actually pushes the middle benzonitrile ring to a position that

allows the cyano group to ﬁt into a narrow cleft between Met-

570 and Tyr-706 in rnNOS (Supporting Information, Figure

S4A). In addition, the cyano group H-bonds with a structural

water molecule in the cleft, which in turn H-bonds with the

protein backbone. In this way, the benzonitrile moiety of those

2-aminoquinoline compounds is tightly sequestered in the

protein cleft, giving high potency. In the case of the 2-

aminopyridine compound with a diamino tail,²⁴the middle

benzonitrile ring points in an opposite direction, close to Ser-

477 in rnNOS (Supporting Information, Figure S4B). That

direction is restrained by the diamine, which tightly interacts

with the heme propionate A. As a result of 3-point H-bonds

To increase the basicity of the amino tail, we prepared 16,

which has almost the same constitution as 15 but with a

methylene group extending between the amino group and the

aromatic biaryl linker. The nNOS inhibitory activity of

analogue 16 (predicted pK_a= 9.49 for the tail amine) was

restored and was comparable to 7 and 8 (Table 1). Crystal

structures of rnNOS-16 (Figure 5A) and hnNOS-16 (Figure

5B) reveal a similar upward binding mode of 6, in which the

middle pyridine ring H-bonds with Tyr562 (rnNOS) or Tyr567

(hnNOS). One distinction exists in the tail amino group: in

rnNOS, the amino group makes a H-bond with a water

molecule bridging between heme propionate A and H₄B,

whereas in hnNOS, the amino group displaces the water

molecule and directly H-bonds with both the heme propionate

and H₄B.

To determine the impact of diﬀerent connectivity to the

inhibitor−enzyme interactions, the position of the

(methylamino)methylene group relative to the pyridine ring

was also changed from meta- for 16 to para- for 17. As shown

in Figure 6, 17 binds to hnNOS in an upward mode with its

middle pyridine H-bonded to Tyr567; however, in rnNOS, the

middle pyridine bends back to interact with heme propionate

D. In both cases, the tail amino group in 17 can no longer

interact with either heme propionate or H₄B because of its

para-position from the pyridine ring. The potency of 17 is

slightly poorer than that of 16 as a result of the lack of this tail

interaction.

f̵rom head, middle, and tail, a very potent inhibitor is obtained.

Therefore, whether or not a functional group can make

favorable contacts (often H-bonds) determines if this group

contributes positively to the potency of the inhibitor.

The next modiﬁcation in our study was to increase the

rigidity, and therefore potentially enhance cell membrane

permeability, of the modiﬁed analogues. Compounds 15−17,

containing a pyridine-based biaryl linker, were then synthesized.

To maintain a similar distance of the amino tail group to the

pyridine ring (via four bonds as found in 6), 15 was prepared

with a methylaniline in the tail. As such, the basicity of the tail

amino group was also modulated (pK_aof 15 = 4.29 predicted

G

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 7. Active site structures of 18 bound to rnNOS (A) and hnNOS (B). The electron density displayed here was calculated by the Polder map

function in PHENIX and contoured at 3.5 σ.

Knowing the structure of 16 and 17, we can comment on

why 15 is a much poorer inhibitor. The tail amino group in 15

is in an aniline, thus having lower basicity. More importantly,

the rigidity and short arm of this amino group keep it from

reaching out to the H₄B site water molecule.

permeability study. The in vitro permeability of selected

compounds was measured using the parallel artiﬁcial membrane

permeability for blood−brain barrier (PAMPA−BBB) assay.²⁸

Additionally, the eﬄux ratio (ER) was determined with a Caco-

2 assay to evaluate their P-gp liability. The PAMPA−BBB assay

was ﬁrst developed by Di et al.²⁸and has been reported to be

one of the most eﬃcient and low-cost assays to evaluate the

BBB permeation of CNS candidates at the early stage of

development.^16,29,30In this assay, porcine brain lipid is used as

an artiﬁcial membrane to predict the passive permeability of

tested compounds. Because the BBB has a tight junction

between endothelial cells, transcellular passive diﬀusion is the

major pathway for CNS drugs to enter the brain.²⁵Five

commercial drugs (Table 2) were used as standard compounds

To obtain analogues with increased cell membrane, and

therefore BBB permeability, we also assessed a combined

modiﬁcation in which both molecular rigidity (as found for

15−17) and lipophilicity (as found for 7−10) were employed

to give 18,19. The chemical structures of 18 and 19 are almost

identical to 16 and 17, respectively, except for the fact that the

pyridine-based biaryl linker is replaced by a ﬂuorophenyl-based

biaryl linker. Biological assays revealed that 18 retains excellent

inhibitory activity against nNOS (18, K_i(rnNOS) = 44 nM, K_i

(hnNOS) = 83 nM), while there is a marked drop in activity for

19 (K_i(rnNOS) = 185 nM, K_i(hnNOS) = 362 nM). This

trend seems to be consistent with biaryl-linker analogues in

which compounds with the (methylamino)methylene group at

the meta-position (16, 18) gave greater potency than those with

a para-substituted (methylamino)methylene (17, 19). Addi-

tionally, there is no signiﬁcant change in nNOS inhibition

between two meta-analogues, while the ﬂuorophenyl-containing

analogue (18) displays more than 2-fold higher selectivity for

hnNOS over heNOS (16, hn/he = 106; 18, hn/he = 234).

Two diﬀerent binding modes were observed for 18 bound to

rnNOS and hnNOS. The middle ﬂuorophenyl ring of 18 bends

back against the heme in rnNOS, as shown in Figure 7A, while

its tail amino group still reaches to the water molecule bridging

between heme propionate A and the H₄B. In contrast, the

ﬂuorophenyl ring of 18 in the hnNOS structure (Figure 7B)

turns upward, making a H-bond with Tyr567 while the tail

amino group folds back to make a H-bond with the H₄B site

water molecule. Either binding mode can achieve fairly good

potency with nNOS. An attempt at getting a heNOS-18

structure was made. The binding mode was found to be similar

to that shown in rnNOS; that is, the middle ﬂuorophenyl ring

bends back against the heme. However, in heNOS, the tail

phenyl ring and amino group are more disordered without

making any favorable contacts with the protein or the water

molecule, which also prevented the reﬁnement completion. In

reference to the structures of rnNOS-17 and hnNOS-17

(Figure 6), it is easy to understand why 19 is a poor inhibitor.

The para-position of the middle ﬂuorophenyl ring makes the

tail amino group impossible to reach the H₄B site water

molecule no matter which binding mode 19 might adopt in

nNOS.

Table 2. Eﬀective Permeability (P_e) of Five Commercial

a

Drugs and nNOS Inhibitors in the PAMPA−BBB Assay

reported P_e

determined P_e

d

b

c

compd

Log D (10⁻⁶cm s⁻¹

)

(10⁻⁶cm s⁻¹

)

prediction

verapamil

16

21.30 1.50

20.52 0.64

8.04 0.41

desipramine

12

chlorpromazine

6.5

dopamine

0.2

0.12 0.011

0.15 0.04

10.4 0.75

14.8 0.69

5.56 0.18

17.41 0.50

theophylline

0.12

7

0.21

1.28

0.59

1.93

CNS (+)

12

16

18

a

All assays were performed over 17 h at a concentration of 200 μM.

b

c

Log D values were predicted using ChemAxon software. Eﬀective

permeability values from literature.²⁸Eﬀective permeability values

d

obtained in our in-house conditions.

to establish and validate our in-house assay. Two drugs,

verapamil and theophylline, were also used as positive and

negative controls, respectively, during each permeability test of

the selected nNOS inhibitors (see Experimental Section for

details). Compared to reported values in the literature (Table

2),²⁸the eﬀective permeability (P_e) values of commercial drugs

obtained under our conditions are slightly higher. Therefore, a

higher cutoﬀ to classify a compound as CNS (+) or CNS (−)

was used. If P_eof a compound is larger than 4.0 × 10⁻⁶cm/s

(compared to a 2.0 × 10⁻⁶cm/s cutoﬀ value in Di’s report),²⁸

the compound was predicted to have good potentialability to

cross the BBB. Table 2 summarizes P_evalues of ﬁve

commercial-drug standards and our selected nNOS inhibitors

(7, 12, 16, and 18). The results reveal that all the selected

Permeability. Compounds 7, 12, 16, and 18, with high

potency and selectivity for nNOS, were selected for a

H

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Table 3. Caco-2 Apparent Permeability and Eﬄux Ratio (ER) of Selected nNOS Inhibitors with Control Compounds

b

apparent permeability (P_app, 10⁻⁶cm s⁻¹

)

recovery

compd

mean A → B

mean B → A

eﬄux ratio

A → B (%)

B → A (%)

f

6

7

0

N.C.

NC

71

73

80

28

NC

96

3.9 0.1

9.2 0.3

0.7 0.1

0.8 0.03

36.1

36.4 1.3

54.2 17.6

47.2 0.8

9.2 2.1

10.1

9.2 0.6

5.9 1.7

70.6 10.3

11.0 3.0

0.28

12

100

82

16

18

c

warfarin

d

ranitidine

0.45

1.2

2.7

e

talinolol

0.1

7.1

71

a

b

c

d

All assays were performed over 2 h at a concentration of 10 μM. Apparent permeability value. High permeability control. Low permeability

e

f

control. High eﬄux control. Not calculated.

nNOS inhibitors exhibit a predicted CNS (+) with P_evalues up

to 17.4 × 10⁻⁶cm/s. Compound 16 (P_e= 5.56 × 10⁻⁶cm/s),

with a pyridine-based biaryl linker, displays the lowest

permeability among the selected compounds, indicating that

the presence of the pyridine ring signiﬁcantly hinders the

permeability of nNOS inhibitors, which is consistent with the

little-to-no permeability found for lead compound 6 in the

Caco-2 assay.

Replacement of the pyridine linker by a ﬂuorobenzene ring

in nNOS inhibitors was found to greatly help increase the P_e

value. For example, 18, containing a ﬂuorophenyl-based biaryl

linker, exhibits a 3-fold higher P_e(17.41 × 10⁻⁶cm/s) than that

of 16 with a pyridine-based biaryl linker (16, P_e= 5.56 × 10⁻⁶

cm/s). Compound 12, with a ﬂuorophenyl linker and a tertiary

amino tail, exhibits a P_eof ca. 1.5-fold higher than that of 7,

which contains a phenyl ring linker and a secondary amino tail.

The PAMPA−BBB assay also supports combining molecular

rigidity and lipophilicity to obtain greater permeability. While

retaining excellent potency and selectivity, compound 18

exhibits a comparable P_eto that of the commercial CNS

drugs verapamil and desipramine.

it still displays a high ER with only 28% recovery in the A → B

direction. This result also suggests that compound 18 is

retained in the lipid layer of colon cells, which is possibly the

result of its high lipophilicity. Obtaining the optimal

lipophilicity is key in CNS drug development to deliver drugs

into the brain; compounds with insuﬃcient lipophilicity will

suﬀer from poor penetration, while compounds with excessive

lipophilicity will have increased risks of toxicity and instability

as well as retention in the cell membrane, preventing it from

crossing the BBB.^25,32Of the four selected compounds, 12

exhibits the lowest ER (5.9) with high recovery percentage in

both directions. This compound also expresses a favorable

permeability in the PAMPA−BBB assay (Table 2). Although its

ER value is still out of the therapeutic range for CNS drugs

(usually ER < 3),³³there has been signiﬁcant improvement in

the permeability as demonstrated by both PAMPA−BBB and

Caco-2 assays compared to lead compound 6. The insights

from these modiﬁcations help to understand further the

structure−activity relationship and cell permeability of nNOS

inhibitors.

Finally, the eﬄux ratio (ER) of the four selected analogues

was determined using a Caco-2 assay to evaluate their P-gp

substrate liability. Eﬄux transporters contribute signiﬁcantly to

limiting the brain penetration of drugs; P-glycoprotein (P-gp)

and breast cancer resistant protein (BCRP) are the two most

important ATP-driven eﬄux transporters at the BBB, of which

P-gp has a larger substrate speciﬁcity and therefore higher

impact than BCRP.¹⁶The bidirectional Caco-2 assay is

commonly used to identify the P-gp liability of a drug,³¹in

which the permeability of a compound is measured by its ability

to cross a monolayer of colon cells with expressed P-gp from

two directions, either from apical to basal (A → B) or from

basal to apical (B → A) wells. The eﬄux ratio (ER) is the ratio

of the apparent permeability (P_app) of B → A over A → B. The

higher the ER a compound has, the less potential to penetrate

the BBB. P_appand ER of the selected nNOS inhibitors (7, 12,

16, and 18) and control compounds are shown in Table 3. The

results indicate that the four selected analogues all exhibit better

permeability than 6. Compound 16, which has a low P_evalue in

the PAMPA−BBB assay (P_e= 5.56 × 10⁻⁶cm/s), also displays

the lowest permeability among the series in the Caco-2 assay

with the highest eﬄux ratio, indicating that this biaryl

containing pyridine compound is not only a poor penetrant

but also a very strong P-gp substrate. The Caco-2 results again

conﬁrm that replacement of the pyridine ring by a

ﬂuorobenzene ring signiﬁcantly increases permeability. Specif-

ically, 18 displays an ER less than one-sixth that of 16, although

CONCLUSION

■

We have designed and synthesized a new series of potent and

selective human nNOS inhibitors based on the 2-aminopyridine

scaﬀold aimed at improving their cell membrane permeability

and therefore increasing their ability to cross the BBB. Diﬀerent

strategies, involving replacing the polar and basic middle

pyridine linker, reducing the number of hydrogen bond donors

and increasing the rigidity of the molecular skeletons, have been

used. We discovered that the combination of exchanging the

middle pyridine linker with a more lipophilic ﬂuorobenzene

linker and the use of a tertiary amine in place of a secondary

amine resulted in compound 12, which not only retains

excellent inhibition for human nNOS (K_i= 30 nM) and a

tremendous selectivity over human eNOS (hn/he = 2799,

which is the highest selectivity ratio we have obtained to date)

but also increases the permeability of the compound compared

to lead compound 6. X-ray diﬀraction analysis of 12 bound to

rat and human nNOS reveals that the ﬂuorine atom of the

ﬂuorobenzene linker forms a H-bonds with Tyr-567 in hnNOS

and Tyr-562 in rnNOS, which mimics what was previously

observed for the nitrogen of the pyridine linker in compound 6.

However, introducing many aromatic rings in 18 resulted in

excessive lipophilicity, which caused retention in the cell

membrane. These results serve as the basis for further

exploration of the properties of these potent and selective

nNOS inhibitors.

I

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

ﬁltrate was concentrated under reduced pressure. The crude product

was subjected to 2,5-dimethylpyrrole deprotection without puriﬁca-

tion. A microwave vial was charged with crude reduction products

26a−c (1 equiv) and NH₂OH·HCl (3−4 equiv). The mixtures were

diluted with EtOH/water (2:1) to form a 0.16 M solution. The

microwave vial was capped, and the reaction mixture was stirred at 100

°C for 20 h. The cap was removed, and the reaction mixture was

concentrated under reduced pressure. The crude product mixture was

puriﬁed by ﬂash chromatography to give 11, 12, and 13.

General Procedure D: Preparation of 32a−c and 34a,b (Suzuki

Cross-Coupling). A microwave vial was charged with Pd(PPh₃)₄(5

mol %), K₂CO₃(2 equiv), boronic acid 31a−c (1.5 equiv), and 30 (1

equiv). The mixture was diluted with THF/water (4:1) to form a 0.2

M solution. The microwave vial was capped, and the reaction mixture

was stirred at 80 °C for 20 h. The cap was removed, and the reaction

mixture was diluted with ethyl acetate. The crude product was ﬁltered,

and the ﬁltrate was dried over Na₂SO₄and concentrated under

reduced pressure to give reaction crude 32a−c and 34a,b.

General Procedure E: Preparation of 33b−c and 35a,b

(Reductive Amination). A scintillation vial was charged with N-

methylamine hydrochloride salt (3 equiv) and NaOAc (3 equiv), and

the mixtures were diluted with MeOH to form a 0.1 M solution. The

resulting slurry was allowed to stir at 25 °C for 15 min. The slurry was

then transferred to a scintillation vial charged with 1 equiv of the crude

product from the Suzuki cross-coupling (32b−c or 34a,b). The

reaction mixtures were then allowed to stir at 25 °C for 1 h. At this

time, Na₂SO₄was added to the scintillation vial, and the mixture was

ﬁltered. The ﬁltrate was cooled to 0 °C followed by addition of NaBH₄

(3 equiv). The reaction mixture was allowed to warm to rt and stirred

for 1 h. The reaction was quenched with water, and the methanol was

removed under reduced pressure. The aqueous mixture was extracted

with DCM three times, and the organic layers were collected, dried

over Na₂SO₄, and concentrated to give the crude product, which was

puriﬁed by ﬂash chromatography to give 33b,c and 35a,b.

General Procedure F: Preparation of 15−19 (Pyrrole Depro-

tection). A microwave vial was charged with 33a−c or 35a−b (1

equiv) and NH₂OH·HCl (3−4 equiv). The mixtures were diluted with

EtOH/water (2:1) to form a 0.16 M solution. The microwave vial was

capped, and the reaction mixture was stirred at 100 °C for 20 h. The

cap was removed, and the reaction mixture was concentrated under

reduced pressure. The crude product mixture was puriﬁed by ﬂash

chromatography to give 15−19.

EXPERIMENTAL SECTION

■

Chemistry. General Procedures. All reagents were purchased from

Sigma-Aldrich and used without further puriﬁcation. Compound 20,

23a, and 30 were prepared according to our previous report.²³Aryl

bromides 21a−d are commercially available. Alkyne 23b and boronic

acids 31a−c were also obtained from commercial sources. Anhydrous

solvents (THF, CH₂Cl₂, DMF) were puriﬁed before use by passing

through a column composed of activated alumina and a supported

copper redox catalyst. Thin layer chromatography (TLC) was

performed on silica gel 60 F254 precoated plates (0.25 mm) from

Silicycle, and components were visualized by ultraviolet light (254 nm)

and/or KMnO₄or ninhydrin stain. Flash column chromatography was

performed on an Agilent 971-FP automated ﬂash puriﬁcation system

with a Varian column station and various Silicycle cartridges (4−80 g,

1

40−63 μm, 60 Å). H and ¹³C NMR spectra were recorded on a

Bruker Avance III NMR spectrometer at 500 and 126 MHz,

respectively, in CDCl₃or CD₃OD. Chemical shifts were reported in

ppm, multiplicities are indicated by s = singlet, d = doublet, t = triplet,

q = quartet, sep = septet, dd = doublet of doublet, dt = doublet of

triplet, m = multiplet, br = broad resonance. Coupling constants J were

reported in Hz. High resolution mass spectral data were obtained on

an Agilent 6210 LC-TOF spectrometer in the positive ion mode using

electrospray ionization with an Agilent G1312A HPLC pump and an

Agilent G1367B autoinjector at the Integrated Molecular Structure

Education and Research Center (IMSERC), Northwestern University.

The purity of compounds was tested by using a reserved-phase

analytical Agilent Inﬁnity 1260 HPLC with an Agilent Poroshell 120

EC-C18 column, detecting with UV absorbance at 254 nm. All

compounds undergoing biological tested had >95% purity. The

preparations of Sonagashira coupling products, Suzuki coupling

products, and ﬁnal compounds are described below, while the

synthesis of intermediates 22c,d is presented in the Supporting

Information.

General Procedure A: Sonagashira Cross Coupling. A microwave

vial was charged with Pd(PPh₃)₄(5 mol %), CuI (5 mol %), and aryl

bromide 22a, 22b, 22c, or 27 (1 equiv). The mixtures were diluted

with triethylamine to form a 0.16 M solution followed by the addition

of alkyne 23a or 23b (1.5−2 equiv). The microwave vial was capped,

and the reaction mixture was stirred at 90 °C for 20 h. The cap was

removed, and the reaction mixture was diluted with ethyl acetate and

ﬁltered. The ﬁltrate was washed with water, ammonium chloride, and

brine, dried with Na₂SO₄, and concentrated under reduced pressure.

The crude product mixture was puriﬁed by ﬂash column

chromatography to give 24a−c, 26a−c, 28a, and 28c.

General Procedure B: Boc Deprotection, Alkyne Reduction, and

Pyrrole Deprotection. Compounds 24a−c (1 equiv) were diluted with

DCM to form a 0.1 M solution followed by addition of TFA (20%

volume). The reaction mixture was allowed to stir at rt for 1 h. At this

time, the crude product was concentrated under reduced pressure,

diluted with CH₂Cl₂, and washed with satd NaHCO₃. The organic

layer was dried over Na₂SO₄and concentrated to give crude products

25a−c. The crude product was used for further steps without

puriﬁcation. A scintillation vial was charged with 10% wt. Pd/C and

crude product 25a−c (1 equiv). The mixtures were diluted with

methanol to form a 0.1 M solution. The reaction mixture was stirred at

rt for 20 h under a hydrogen balloon (1 atm). At this time, the crude

product was ﬁltered, and the ﬁltrate was concentrated under reduced

pressure to give a crude product. A microwave vial was charged with

reduction crude product (1 equiv) and NH₂OH·HCl (3−4 equiv).

The mixtures were diluted with EtOH/water (2:1) to form a 0.16 M

solution. The microwave vial was then capped, and the reaction

mixture was stirred at 100 °C for 20 h. The cap was removed, and the

reaction mixture was concentrated under reduced pressure. The crude

product mixture was puriﬁed by ﬂash chromatography to give 7, 8, and

9.

4-Methyl-6-(3-(3-(methylamino)propyl)phenethyl)pyridin-2-

amine (7). Compound 7 was synthesized according to general

procedure B using 24a (130.0 mg, 0.284 mmol), TFA (0.5 mL), 10%

wt Pd/C (10.8 mg), and NH₂OH·HCl (59.1 mg). 7 was isolated as a

brown oil (55.5 mg, 69%) after ﬂash column chromatography

(MeOH:DCM 3:7). ¹H NMR (500 MHz, methanol-d₄): δ 7.20−

7.18 (m, 2H), 7.10−7.08 (m, 2H), 6.71 (s, 1H), 6.59 (s, 1H), 3.03−

3.00 (m, 6H), 2.71−2.69 (m, 5H), 2.32 (s, 3H), 2.08−1.97 (m, 2H).

¹³C NMR (125 MHz, methanol-d₄): δ 157.6, 154.3, 148.6, 140.7,

140.0, 128.5, 128.4, 126.4, 126.3, 113.8, 109.5, 48.8, 48.7, 34.5, 34.4,

32.1, 27.6, 20.9. HRMS ESI: calcd for C₁₈H₂₆N₃[M + H]⁺284.2121,

found 284.2121.

6-(3-Fluoro-5-(3-(methylamino)propyl)phenethyl)-4-methylpyri-

din-2-amine (8). Compound 8 was synthesized according to general

procedure B using 24b (135.0 mg, 0.284 mmol), TFA (0.58 mL), 10%

wt Pd/C (17.0 mg), and NH₂OH·HCl (64.0 mg). 8 was isolated as a

brown oil (74.5 mg, 87.07%) after ﬂash column chromatography

1

(MeOH:DCM= 3:7). H NMR (500 MHz, methanol-d₄): δ 7.02 (s,

1H), 6.89−6.85 (m, 2H), 6.73 (s, 1H), 6.62 (s, 1H), 3.04−3.01 (m,

6H), 2.73−2.70 (m, 5H), 2.33 (s, 3H), 2.06−2.00 (m, 2H). ¹³C NMR

(125 MHz, methanol-d₄): δ 163.0 (d, J_C−F= 243.0 Hz), 157.6, 154.4,

148.3, 143.4 (d, J_C−F= 7.8 Hz), 142.6 (d, J_C−F= 7.8 Hz), 124.2 (d,

J_C−F= 2.1 Hz), 113.7, 112.9 (d, J_C−F= 21.3 Hz), 112.8 (d, J_C−F= 21.4

Hz), 109.5, 48.5, 34.1, 34.0, 32.4, 31.8, 27.3, 20.6. HRMS ESI: calcd for

C₁₈H₂₅FN₃[M + H]⁺302.2027, found 302.2031.

4-Methyl-6-(3-(3-(methylamino)propyl)-5-(triﬂuoromethyl)-

phenethyl)pyridin-2-amine (9). Compound 9 was synthesized

according to general procedure B using 24c (222 mg, 0.42 mmol),

General Procedure C: Alkyne Reduction and Pyrrole Depro-

tection. A scintillation vial was charged with 10% wt. Pd/C and 26a−c

(1 equiv). The mixtures were diluted with methanol to form a 0.1 M

solution. The reaction mixture was stirred at rt for 20 h under a

hydrogen balloon (1 atm), then the crude product was ﬁltered and the

J

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

TFA (0.8 mL), 10% wt Pd/C (16.1 mg), and NH₂OH·HCl (72 mg). 9

was isolated as a brown oil (80.2 mg, 54%) after ﬂash column

chromatography (MeOH:DCM 1:3). ¹H NMR (500 MHz, methanol-

d₄): δ 7.54 (s, 1H), 7.43 (s, 1H), 7.40 (s, 1H), 6.74 (s, 1H), 6.62 (s,

1H), 3.16−3.12 (m, 2H), 3.09−3.06 (m, 4H), 2.83 (t, J = 8.0 Hz, 2H),

2.74 (s, 3H), 2.34 (s, 3H), 2.11−2.05 (m, 2H). ¹³C NMR (125 MHz,

methanol-d₄): δ 157.5, 154.5, 148.2, 142.3, 141.3, 132.3, 130.6 (q, J_C−F

= 32.5 Hz), 124.3 (q, J_C−F= 270 Hz), 122.9 (d, J_C−F= 3 Hz), 122.8 (d,

J_C−F= 3 Hz), 113.8, 109.6, 48.5, 34.1, 34.0, 32.4, 31.8, 27.3, 20.6.

HRMS ESI: calcd for C₁₉H₂₄F₃N₃[M + H]⁺352.1995; found

352.2011.

3-(2-(6-Amino-4-methylpyridin-2-yl)ethyl)-5-(3-(methylamino)-

propyl)benzonitrile (10). Compound 10 was synthesized by Boc

deprotection of 28a followed by pyrrole deprotection, and alkyne

reduction. 10 (182.7 mg, 0.38 mmol) was diluted with DCM to form a

0.1 M solution followed by addition of TFA (20% volume). The

reaction mixture was allowed to stir at rt for 1 h. At this time, the crude

product was concentrated under reduced pressure, diluted with DCM,

and washed with satd NaHCO₃. The organic layer was dried over

MgSO₄and concentrated to give 28b. The crude product was used for

further steps without puriﬁcation. 28b was subjected to pyrrole

deprotection according to general procedure B using NH₂OH·HCl

(82 mg) and EtOH/water (2/1, 2.1 mL). 29a was isolated as a light-

yellow oil (49.3 mg, 43%) after ﬂash column chromatography

(MeOH:DCM 1:3). A scintillation vial was charged with 10% wt

Pd/C (4.6 mg) and 29a (49.3 mg). The mixtures were diluted with

methanol to form a 0.1 M solution. The reaction mixture was stirred at

rt for 20 h under hydrogen gas, then ﬁltered, and the ﬁltrate was

concentrated under reduced pressure to give 10. ¹H NMR (500 MHz,

methanol-d₄): δ 7.64 (s, 2H), 7.52 (s, 1H), 7.39 (s, 1H), 6.71 (s, 1H),

6.61 (s, 1H), 3.08−3.01 (m, 8H), 2.79−2.73 (m, 2H), 2.71 (s, 3H),

2.32 (s, 3H), 2.07−2.04 (m, 2H). ¹³C NMR (125 MHz, methanol-d₄):

δ 157.5, 157.4, 154.5, 148.5, 141.3, 140.6, 134.0, 132.0, 125.6, 125.5,

113.7, 109.5, 48.5, 34.3, 34.1, 32.4, 31.8, 27.3, 20.6. HRMS ESI: calcd

for C₁₉H₂₄N₄[M + H]⁺309.2073, found 309.2081.

2H). ¹³C NMR (125 MHz, methanol-d₄): δ 157.6, 154.4, 148.2, 142.1,

141.4, 132.3l, 130.6 (q, J = 31.7 Hz), 124.0 (q, J = 271.6 Hz), 123.1

(m, 2C), 113.8, 109.5, 57.0, 42.2, 34.1, 34.0, 31.7, 25.8, 20.6. HRMS

ESI: calcd for C₂₀H₂₆F₃N₃[M + H]⁺366.2151, found 366.2163.

3-(2-(6-Amino-4-methylpyridin-2-yl)ethyl)-5-(3-(dimethylamino)-

propyl)benzonitrile (14). Compound 14 was synthesized from 28c by

ﬁrst removing 2,5-dimethylpyrrole protecting group using NH₂OH·

HCl in EtOH/H₂O (2:1). Product 29b was then subjected to

hydrogenation with 10% wt Pd/C in MeOH (0.1 M). The reaction

mixture was stirred at rt for 20 h under hydrogen gas. After that, the

crude was ﬁltered and the ﬁltrate was concentrated under reduced

pressure to give a crude 14 which was isolated as a brown oil after ﬂash

1

column chromatography (MeOH:DCM 3:7). H NMR (500 MHz,

methanol-d₄): δ 7.61 (s, 1H), 7.60 (s, 1H), 7.28 (s, 1H), 6.35 (s, 1H),

6.32 (s, 1H), 3.04−2.98 (m, 2H), 2.96−2.91 (m, 2H), 2.90−2.84 (m,

2H), 2.76−2.70 (m, 8H), 2.20 (s, 3H), 2.05−1.95 (m, 2H). ¹³C NMR

(125 MHz, methanol-d₄): δ 171.1, 158.9, 157.2, 150.4, 142.2, 141.1,

133.9, 131.9, 125.3, 125.1, 113.3, 106.8, 57.5, 42.6, 38.4, 35.5, 32.1,

26.6, 19.7. HRMS ESI: calcd for C₂₀H₂₆N₄[M + H]⁺323.2230, found

323.2251.

4-Methyl-6-(2-(5-(3-(methylamino)phenyl)pyridin-3-yl)ethyl)-

pyridin-2-amine (15). Compound 15 was synthesized according to

general procedure F using 33a (139 mg, 0.35 mmol) and NH₂OH·

HCl (86 mg). 15 was isolated as a light-yellow oil (75.7 mg, 68%) after

1

ﬂash column chromatography (methanol:DCM 1:4). H NMR (500

MHz, methanol-d₄): δ 8.66 (s, 1H), 8.46 (s, 1H), 8.02 (s, 1H), 7.22 (t,

J = 7.5 Hz, 1H), 7.89−6.87 (m, 2H), 6.71 (s, 1H), 6.69 (d, J = 7.5 Hz,

1H), 6.62 (s, 1H), 3.18−3.15 (m, 2H), 3.10−3.07 (m, 2H), 2.82 (s,

3H), 2.31 (s, 3H). ¹³C NMR (125 MHz, methanol-d₄): δ 157.5, 154.5,

150.6, 148.0, 146.7, 144.7, 138.1, 137.4, 136.2,, 135.7, 129.5, 115.2,

113.8, 112.5, 110.6, 109.7, 33.8, 31.4, 29.5, 20.6. HRMS ESI: calcd for

C₂₀H₂₃N₄[M + H]⁺319.1924, found 319.1917.

4-Methyl-6-(2-(5-(3-((methylamino)methyl)phenyl)pyridin-3-yl)-

ethyl)pyridin-2-amine (16). Compound 16 was synthesized according

to general procedure F using 33b (98.2 mg, 0.24 mmol) and NH₂OH·

HCl (53 mg). 16 was isolated as a light-yellow oil (40.2 mg, 51%) after

6-(3-(3-(Dimethylamino)propyl)phenethyl)-4-methylpyridin-2-

amine (11). Compound 11 was synthesized according to general

procedure C using 26a (89.4 mg, 0.2408 mmol), 10% wt Pd/C (12.3

mg), and NH₂OH·HCl (53.2 mg). 11 was isolated as a brown oil (57.9

1

ﬂash column chromatography (methanol:DCM 1:4). H NMR (500

MHz, methanol-d₄): δ 8.79 (s, 1H), 8.51 (s, 1H), 8.30 (s, 1H), 8.01 (s,

1H), 7.81 (t, J = 6.0 Hz, 1H), 7.61 (d, J = 6.0 Hz, 2H), 6.70 (s, 1H),

6.69 (s, 1H), 4.33 (s, 2H), 3.23−3.20 (m, 2H), 3.15−3.13 (m, 2H),

2.79 (s, 3H), 2.34 (s, 3H). ¹³C NMR (125 MHz, methanol-d₄): δ

157.7, 154.5, 147.9, 147.1, 144.5, 137.5, 136.8, 136.5, 136.4, 132.4,

129.8, 129.7, 128.6, 127.9, 113.8, 109.7, 52.0, 33.6, 31.9, 31.3, 20.6.

HRMS ESI: calcd for C₂₁H₂₅N₄[M + H]⁺333.2074, found 333.2081.

4-Methyl-6-(2-(5-(4-((methylamino)methyl)phenyl)pyridin-3-yl)-

ethyl)pyridin-2-amine (17). Compound 17 was synthesized according

to general procedure F using 33c (109.7 mg, 0.27 mmol) and

NH₂OH·HCl (59 mg). 17 was isolated as a light-yellow oil (85.2 mg,

1

mg, 81%) after ﬂash column chromatography (MeOH:DCM 3:7). H

NMR (500 MHz, methanol-d₄): δ 7.22 (t, J = 7.5 Hz, 1H), 7.19 (s,

1H), 7.12−7.08 (m, 2H), 6.68 (s, 1H), 6.59 (s, 1H), 3.17−3.14 (m,

2H), 3.03−3.01 (m, 4H), 2.89 (s, 6H), 2.71−2.68 (m, 2H), 2.33 (s,

3H), 2.09−2.03 (m, 2H). ¹³C NMR (125 MHz, methanol-d₄): δ 157.6,

154.4, 148.7, 140.5, 140.0, 128.5, 128.3, 126.4, 126.2, 113.7, 109.4,

57.2, 42.1, 34.4, 34.3, 31.9, 26.0, 20.6. HRMS ESI: calcd for C₁₉H₂₈N₃

[M + H]⁺298.2278, found 298.2279.

6-(3-(3-(Dimethylamino)propyl)-5-ﬂuorophenethyl)-4-methylpyr-

idin-2-amine (12). Compound 12 was synthesized according to

general procedure C using 26b (189.5 mg, 0.4869 mmol), 10% wt Pd/

C (20.1 mg), and NH₂OH·HCl (94.1 mg). Compound 12 was

isolated as a brown oil (118.6 mg, 77%) after ﬂash column

chromatography (MeOH:DCM 3:7). ¹H NMR (500 MHz, meth-

anol-d₄): δ 7.04 (s, 1H), 6.93 (d, J = 9.5 Hz, 2H), 6.71 (s, 1H), 6.62 (s,

1H), 3.81 (dd, J = 8.0, 5.0 Hz, 2H), 3.04−3.02 (m, 4H), 2.91 (s, 6H),

2.71 (dd, J = 8.0, 7.0 Hz, 2H), 2.33 (s, 3H), 2.11−2.04 (m, 2H). ¹³C

NMR (125 MHz, methanol-d₄): δ 163.0 (d, J_C−F= 243.0 Hz), 157.3,

154.4, 148.4, 143.3 (d, J_C−F= 7.8 Hz), 142.7 (d, J_C−F= 7.9 Hz), 124.2

1

95%) after ﬂash column chromatography (methanol:DCM 1:3). H

NMR (500 MHz, methanol-d₄): δ 8.70 (s, 1H), 8.46 (s, 1H), 8.12 (s,

1H), 7.80 (d, J = 8.0 Hz, 2H), 7.68 (d, J = 8.0 Hz, 2H), 6.69 (s, 1H),

6.65 (s, 1H), 5.87 (s, 2H), 4.28 (s, 2H), 3.20−3.17 (m, 2H), 3.13−

3.10 (m, 2H), 2.77 (s, 3H), 2.33 (s, 3H). ¹³C NMR (125 MHz,

methanol-d₄): δ 157.5, 154.6, 148.1, 147.9, 145.2, 138.2, 136.4, 136.1,

135.5, 131.4, 130.5, 127.6, 113.7, 109.6, 51.7, 33.7, 31.9, 31.4, 20.5.

HRMS ESI: calcd for C₂₁H₂₅N₄[M + H]⁺333.2074, found 333.2070.

6-(2-(5-Fluoro-3′-((methylamino)methyl)-[1,1′-biphenyl]-3-yl)-

ethyl)-4-methylpyridin-2-amine (18). Compound 18 was synthesized

according to general procedure F using 35a (141.9 mg, 0.33 mmol)

and NH₂OH·HCl (72 mg). 18 was isolated as a brown-yellow oil

(92.8 mg, 80%) after ﬂash column chromatography (methanol:DCM

(d, J_C−F= 2.1 Hz), 113.7, 113.0 (d, J_C−F= 21.3 Hz), 112.9 (d, J_C−F

=

21.4 Hz), 109.5, 57.0, 42.2, 34.1, 34.0, 31.7, 25.7, 20.6. HRMS ESI:

calcd for C₁₉H₂₇FN₃[M + H]⁺316.2184, found 316.2192.

1

1:4). H NMR (500 MHz, methanol-d₄): δ 7.90 (s, 1H), 7.72 (d, J =

6-(3-(3-(Dimethylamino)propyl)-5-(triﬂuoromethyl)phenethyl)-4-

methylpyridin-2-amine (13). Compound 13 was synthesized

according to general procedure C using 26c (146 mg, 0.3 mmol),

10% wt Pd/C (16 mg), and NH₂OH·HCl (63 mg). Compound 13

was isolated as a brown oil (128 mg, 85%) after ﬂash column

chromatography (MeOH:DCM 3:7). ¹H NMR (500 MHz, methanol-

d₄): δ 7.56 (s, 1H), 7.46 (s, 1H), 7.41 (s, 1H), 6.74 (s, 1H), 6.64 (s,

1H), 3.28−3.20 (m, 2H), 3.18−3.12 (m, 2H), 3.11−3.05 (m, 2H),

2.94 (s, 6H), 2.82 (t, J = 7.9 Hz, 2H), 2.35 (s, 3H), 2.18−2.08 (m,

7.5 Hz, 1H), 7.57−7.50 (m, 3H), 7.30 (d, J = 9.5 Hz, 1H), 7.05 (d, J =

9.5 Hz, 1H), 6.67 (s, 1H), 6.65 (s, 1H), 4.29 (s, 2H), 3.15−3.12 (m,

2H), 3.11−3.08 (m, 2H), 2.77 (s, 3H), 2.33 (s, 3H). ¹³C NMR (125

MHz, methanol-d₄): δ 163.4 (d, J_C−F= 243.8 Hz), 157.7, 154.4, 148.3,

143.1 (d, J_C−F= 7.5 Hz), 142.5 (d, J_C−F= 8.8 Hz), 140.5, 132.0, 129.5,

129.0, 128.4, 127.8, 123.0, 114.2 (d, J_C−F= 21.3 Hz), 113.7, 111.5 (d,

J_C−F= 22.5 Hz), 109.5, 52.1, 34.2, 34.0, 31.9, 20.6. HRMS ESI: calcd

for C₂₂H₂₅FN₃[M + H]⁺350.2030, found 350.2030.

K

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

6-(2-(5-Fluoro-4′-((methylamino)methyl)-[1,1′-biphenyl]-3-yl)-

ethyl)-4-methylpyridin-2-amine (19). Compound 19 was synthesized

according to general procedure F using 35b (127.5 mg, 0.30 mmol)

and NH₂OH·HCl (65 mg). 19 was isolated as a brown-yellow oil (89

mg, 85%) after ﬂash column chromatography (methanol:DCM 1:4).

¹H NMR (500 MHz, methanol-d₄): δ 7.71 (d, J = 8.0 Hz, 2H), 7.62

(d, J = 8.0 Hz, 2H), 7.40 (s, 1H), 7.23 (d, J = 9.5 Hz, 1H), 7.06 (d, J =

9.5 Hz, 1H), 6.68 (s, 1H), 6.62 (s, 1H), 4.26 (s, 2H), 3.14−3.11 (m,

2H), 3.08−3.05 (m, 2H), 2.76 (s, 3H), 2.32 (s, 3H). ¹³C NMR (125

MHz, methanol-d₄): δ 161.8 (d, J_C−F= 243.8 Hz), 156.1, 152.9, 146.7,

141.6 (d, J_C−F= 7.5 Hz), 140.9 (d, J_C−F= 8.8 Hz), 139.2 (d, J_C−F= 1.3

3-(3-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)-5-ﬂuorophenyl)-N,N-dimethylprop-2-yn-1-amine (26b).

Compound 26b was synthesized according to general procedure A

using Pd(PPh₃)₄(28.9 mg), CuI (5.1 mg), alkyne 23b (125.8 mg), and

22b (193.0 mg, 0.50 mmol). 26b was isolated as a brown oil (189.5

mg, 97%) after ﬂash column chromatography (methanol:DCM 1:19).

¹H NMR (500 MHz, CDCl₃): δ 7.03 (s, 1H), 6.93 (d, J = 9.0 Hz, 1H),

6.87 (s, 1H), 6.84 (s, 1H), 6.80 (d, J = 9.5 Hz, 1H), 5.87 (s, 2H), 3.43

(s, 2H), 3.04−3.02 (m, 4H), 2.35 (s, 3H), 2.34 (s, 6H), 2.10 (s, 6H).

¹³C NMR (125 MHz, CDCl₃): δ 162.4 (d, J_C−F= 244.4 Hz), 160.2,

151.8, 149.6, 144.1 (d, J_C−F= 8.0 Hz), 128.5, 127.8 (d, J_C−F= 2.4 Hz),

124.7 (d, J_C−F= 10.0 Hz), 122.7, 120.3, 116.1 (d, J_C−F= 22.8 Hz),

115.5 (d, J_C−F= 21.0 Hz), 106.8, 85.4, 84.3 (d, J_C−F= 3.4 Hz), 48.5,

44.3, 39.1, 35.1, 21.0, 13.2. MS ESI [M + H]⁺= 390.23

3-(3-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)-5-(triﬂuoromethyl)phenyl)-N,N-dimethylprop-2-yn-1-amine

(26c). Compound 26c was synthesized according to general procedure

A using Pd(PPh₃)₄(28.8 mg), CuI (5 mg), alkyne 23b (128.9 mg),

and 22c (218 mg, 0.50 mmol). 26c was isolated as a brown oil (132

mg, 60%) after ﬂash column chromatography (methanol:DCM 1:19).

¹H NMR (500 MHz, CDCl₃): δ 7.32 (s, 1H), 7.24 (d, J = 1.5 Hz, 1H),

7.09 (d, J = 2.0 Hz, 1H), 6.68 (s, 1H), 6.67 (s, 1H), 5.70 (s, 2H), 3.28

(s, 2H), 2.97−2.84 (m, 4H), 2.18 (s, 6H), 2.17 (s, 3H), 1.93 (s, 6H).

¹³C NMR (125 MHz, CDCl₃): δ δ 171.1, 159.9, 151.8, 149.6, 142.6,

134.9, 130.8 (q, J_C−F= 32.3 Hz), 128.4, 126.2 (q, J_C−F= 3.8 Hz),

124.85 (q, J_C−F= 3.8 Hz), 124.8, 124.1, 122.8, 120.4, 106.8, 86.1, 84.0,

48.5, 44.3, 39.1, 35.1, 20.9, 13.2.

tert-Butyl N-(3-(3-Cyano-5-(2-(6-(2,5-dimethyl-1H-pyrrol-1-yl)-4-

methylpyridin-2-yl)ethyl)phenyl)prop-2-yn-1-yl)(methyl)carbamate

(28a). Compound 28a was synthesized according to general procedure

A using Pd(PPh₃)₄(28.8 mg), CuI (4.0 mg), alkyne 23a (112 mg),

and 27 (191 mg, 0.5 mmol). 28a was isolated as a brown oil (183.7

mg, 75%) after ﬂash column chromatography (MeOH:DCM 1:19).

¹H NMR (500 MHz, CDCl₃): δ 7.47 (s, 1H), 7.42 (s, 1H), 7.33 (s,

1H), 6.86 (s, 1H), 6.85 (s, 1H), 5.86 (s, 2H), 4.24 (brs, 2H), 3.08−

3.02 (m, 4H), 2.93 (s, 3H), 2.34 (s, 3H), 2.08 (s, 6H), 1.46 (s, 9H).

¹³C NMR (125 MHz, CDCl₃): δ 159.5, 155.3, 151.8, 149.8, 143.2,

Hz), 129.4, 128.8, 125.9, 121.3 (d, J_C−F= 2.5 Hz), 112.7 (d, J_C−F

=

21.3 Hz), 112.2, 110.0 (d, J_C−F= 22.5 Hz), 108.0, 50.3, 32.7, 32.5,

30.3, 19.1. HRMS ESI: calcd for C₂₂H₂₅FN₃[M + H]⁺350.2030,

found 350.2030.

tert-Butyl-3-(3-(2-(6-(2,5-dimethyl-1H-pyrrol-1-yl)-4-methylpyri-

din-2-yl)ethyl)phenyl)prop-2-ynyl(methyl)carbamate (24a). Com-

pound 24a was synthesized according to general procedure A using

Pd(PPh₃)₄(28.8 mg), CuI (4.7 mg), alkyne 23a (126.8 mg, 0.75

mmol), and 22a (184.3 mg, 0.50 mmol). 24a was isolated as a yellow

oil (70.2 mg, 31%) after ﬂash column chromatography (ethyl

acetate:hexanes 1:4). ¹H NMR (500 MHz, CDCl₃): δ 7.24−7.21

(m, 2H), 7.17 (t, J = 7.5 Hz, 1H), 7.09 (d, J = 7.5 Hz, 1H), 6.87 (s,

1H), 6.83 (s, 1H), 5.87 (s, 2H), 4.25 (brs, 2H), 3.05−3.02 (m, 4H),

2.95 (s, 3H), 2.34 (s, 3H), 2.10 (s, 6H), 1.47 (s, 9H). ¹³C NMR (125

MHz, CDCl₃): δ 160.6, 155.3, 151.7, 149.5, 141.6, 131.8, 129.4, 128.6,

128.5, 128.3, 122.8, 122.6, 120.2, 106.7, 84.4, 83.7, 80.1, 39.5, 35.5,

33.5, 28.4, 28.3, 21.0, 13.2.

tert-Butyl-3-(3-(2-(6-(2,5-dimethyl-1H-pyrrol-1-yl)-4-methylpyri-

din-2-yl)ethyl)-5-ﬂuorophenyl)prop-2-ynyl(methyl)carbamate

(24b). Compound 24b was synthesized according to general

procedure A using Pd(PPh₃)₄(28.8 mg), CuI (4.9 mg), alkyne 23a

(125.6 mg), and 22b (193.1 mg, 0.50 mmol). 24b was isolated as a

yellow oil (139.2 mg, 60%) after ﬂash column chromatography (ethyl

1

acetate:hexanes 3:7). H NMR (500 MHz, CDCl₃): δ 7.02 (s, 1H),

6.91 (d, J = 9.0 Hz, 1H), 6.88 (s, 1H), 6.84 (s, 1H), 6.81 (d, J = 9.5

Hz, 1H), 5.86 (s, 2H), 4.24 (brs, 2H), 3.04−3.02 (m, 4H), 2.94 (s,

3H), 2.35 (s, 3H), 2.10 (s, 6H), 1.47 (s, 9H). ¹³C NMR (125 MHz,

CDCl₃): δ 162.4 (d, J_C−F= 244.5 Hz), 160.1, 155.3, 151.7, 149.6,

144.1 (d, J_C−F= 7.5 Hz), 128.5, 127.8 (d, J_C−F= 2.4 Hz), 124.4 (d,

J_C−F= 10.0 Hz), 122.7, 120.3, 116.1 (d, J_C−F= 22.9 Hz), 115.8 (d, J_C−F

= 21.0 Hz), 106.7, 85.4, 82.5 (d, J_C−F= 3.4 Hz), 80.2, 39.0, 35.1, 35.0,

33.6, 28.4, 21.0, 13.2. MS ESI [M + Na]⁺= 498.21.

136.1, 132.7, 131.5, 128.4, 124.4, 122.7, 120.5, 118.1, 112.7, 106.8,

87.1, 81.4, 80.3, 38.8, 34.8, 34.7, 33.7, 28.4, 20.9, 13.2. MS ESI [M +

Na]⁺= 505.21

3-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)ethyl)-

5-(3-(dimethylamino)prop-1-yn-1-yl)benzonitrile (28c). Compound

28c was synthesized according to general procedure A using

Pd(PPh₃)₄(28.8 mg), CuI (4.0 mg), alkyne 23b (128.9 mg), and

27 (191 mg, 0.5 mmol). 28c was isolated as a brown oil (148 mg,

75%) after ﬂash column chromatography (MeOH:DCM 1:19). ¹H

tert-Butyl(3-(3-(2-(6-(2,5-dimethyl-1H-pyrrol-1-yl)-4-methylpyri-

din-2-yl)ethyl)-5-(triﬂuoromethyl)phenyl)prop-2-yn-1-yl)(methyl)-

carbamate (24c). Compound 24c was synthesized according to

general procedure A using Pd(PPh₃)₄(29.1 mg), CuI (5.0 mg), alkyne

23a (131.0 mg), and 22c (198.1 mg, 0.50 mmol). 24c was isolated as a

yellow oil (222 mg, 85%) after ﬂash column chromatography (ethyl

1

NMR (500 MHz, CDCl₃) δ, H NMR (500 MHz, chloroform-d) δ

7.30 (s, 1H), 7.24 (s, 1H), 7.14 (s, 1H), 6.68 (s, 1H), 6.66 (s, 1H),

5.67 (s, 2H), 3.24 (s, 2H), 2.93−2.80 (m, 4H), 2.16 (s, 3H), 2.14 (s,

6H), 1.89 (s, 6H). ¹³C NMR (125 MHz, CDCl₃): δ 159.6, 151.8,

149.8, 143.2, 136.1, 132.7, 131.3, 128.4, 124.7, 122.7, 120.5, 118.2,

112.6, 106.8, 87.2, 83.2, 48.5, 44.3, 38.8, 34.7, 21.0, 13.2.

1

acetate:hexanes 1:4). H NMR (500 MHz, CDCl₃): δ 7.48 (s, 1H),

7.41 (s, 1H), 7.27 (s, 1H), 6.86 (s, 2H), 5.88 (s, 2H), 4.26 (brs, 2H),

3.12−3.09 (m, 2H), 3.08−3.05 (m, 2H), 2.97 (s, 3H), 2.35 (s, 3H),

2.10 (s, 6H), 1.48 (s, 9H). ¹³C NMR (125 MHz, CDCl₃): δ 159.8,

155.3, 151.8, 149.7, 142.7, 135.0, 130.8 (q, J_C−F= 32.5 Hz), 128.4,

126.2 (d, J_C−F= 3.1 Hz), 125.1 (d, J_C−F= 3.8 Hz), 123.8, 123.7 (q,

J_C−F= 271.3 Hz), 122.8, 120.4, 106.8, 86.1, 82.2, 80.2, 39.1, 35.1, 33.6,

28.4, 20.9, 13.2. MS ESI [M + Na]⁺= 548.20.

3-(3-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)phenyl)-N,N-dimethylprop-2-yn-1-amine (26a). Compound

26a was synthesized according to general procedure A using

Pd(PPh₃)₄(28.8 mg), CuI (5.0 mg), alkyne 23b (128.9 mg), and

22a (184.4 mg, 0.50 mmol). 26a was isolated as a brown oil (89.4 mg,

48%) after ﬂash column chromatography (ethyl acetate:hexanes 6:4).

¹H NMR (500 MHz, CDCl₃): δ 7.26−7.24 (m, 2H), 7.17 (t, J = 7.5

3-(5-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)pyridin-3-yl)-N-methylaniline (33a). Compound 33a was

prepared via Suzuki cross-coupling and Boc deprotection. 32a was

synthesized according to general procedure D using Pd(PPh₃)₄(24.1

mg), K₂CO₃(141 mg), boronic acid 31a (172.6 mg), and 30 (149 mg,

0.4 mmol). Reaction product 32a was diluted with DCM to form a 0.1

M solution followed by addition of TFA (20% volume). The reaction

mixture was allowed to stir at rt for 1 h. At this time, the crude product

was concentrated under reduced pressure, diluted with DCM, and

washed with satd NaHCO₃. The organic layer was dried over Na₂SO₄

and concentrated to give crude product. 33a was isolated as a light-

yellow oil (139 mg, 70%) after ﬂash column chromatography

(methanol:DCM 1:19). ¹H NMR (500 MHz, CDCl₃): δ 8.65 (s,

1H), 8.37 (s, 1H), 7.25 (t, J = 7.5 Hz, 1H), 7.65 (s, 1H), 6.90 (s, 1H),

6.86 (s, 1H), 6.83 (d, J = 7.5 Hz, 1H), 6.71 (s, 1H), 6.62 (d, J = 7.5

Hz, 1H), 5.88 (s, 2H), 3.94 (brs, 1H), 3.14−3.12 (m, 4H), 2.87 (s,

3H), 2.34 (s, 3H), 2.10 (s, 6H). ¹³C NMR (125 MHz, CDCl₃): δ

160.1, 151.8, 149.9, 149.7, 148.3, 146.0, 138.8, 136.9, 136.6, 134.7,

Hz, 1H), 7.09 (d, J = 7.5 Hz, 1H), 6.87 (s, 1H), 6.84 (s, 1H), 5.87 (s,

2H), 3.44 (s, 2H), 3.06−3.01 (m, 4H), 2.35 (s, 6H), 2.34 (s, 3H), 2.10

(s, 6H). ¹³C NMR (125 MHz, CDCl₃): δ 160.7, 151.7, 149.5, 141.6,

131.8, 129.3, 128.5, 128.4, 128.3, 123.2, 122.7, 120.1, 106.7, 85.4, 84.4,

48.6, 44.3, 39.5, 35.5, 21.0, 13.2.

L

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

128.6, 128.5, 122.8, 120.4, 116.1, 112.2, 110.8, 106.8, 39.2, 32.7, 30.7,

21.0, 13.3. MS ESI [M + Na]⁺= 419.26.

1.3 Hz), 122.7, 120.3, 114.1 (d, J_C−F= 21.3 Hz), 111.5 (d, J_C−F= 21.3

Hz), 106.8, 55.0, 39.4, 35.6, 35.3, 21.0, 13.2.

NOS Enzyme Inhibition Assay. The NOS inhibitory activity of

the studied compounds was measured by the hemoglobin (Hb) NO

capture assay, in which NO production was monitored by a rapid

oxidation of oxyHb to metHb by NO.²⁶Puriﬁed NOSs, including rat

nNOS, human nNOS, murine macrophage iNOS, and human eNOS

used in this study are recombinant enzymes. They are expressed in

Escherichia coli and puriﬁed as previously reported.³⁴⁻³⁷The assay was

done in 100 mM HEPES buﬀer with 10% glycerol (pH 7.4) at 37 °C

in the presence of 10 μM ^L-Arg, 10 μM H₄B, 100 μM NADPH, 0.83

mM CaCl₂, 320 units/mL of calmodulin, and 3 μM human

oxyhemoglobin. The concentration of ^L-Arg, 10 μM, was used as it

is suﬃcient not to cause NOS uncoupling and is close to the K_mvalues

of all three NOS isoforms where competitive inhibitors can be

detected eﬀectively. In the case of iNOS, CaCl₂and calmodulin were

omitted and replaced by HEPES buﬀer because iNOS activation is

calcium-independent. The assay was performed in 96-well plates using

a Biotek Gen5 microplate reader. The NOS enzymes and hemoglobin

were dispensed automatically by the plate reader. NO production was

kinetically monitored at 401 nm for 6 min. The inhibition constants

(K_i) for all NOSs were calculated from the IC₅₀values of the dose−

response curves using the Cheng−Prusoﬀ equation: K_i= IC₅₀/(1 +

[S]/Km)²⁷and K_m(human nNOS, 1.6 μM; rat nNOS, 1.3 μM;

murine iNOS, 8.2 μM; bovine eNOS, 1.7 μM; human eNOS, 3.9

μM).³⁸Dose−response curves were constructed from seven to nine

test concentrations (200 μM to 50 nM), and IC₅₀values were

calculated by nonlinear regression using GraphPad Prism software.

The calculated standard deviations from dose−response curves of the

assays were less than 10% with all NOSs. To conﬁrm that the change

in the absorbance of hemoglobin is caused by the released NO without

any interference from possible binding between the 2-aminopyridine

moiety in 6−19 and the heme in hemoglobin, a control experiment

was performed by monitoring the UV−vis absorption of hemoglobin

without and in the presence of compound 12. The result (Supporting

Information, Figure S5) revealed that there is no change in either λ_max

or the absorbance intensity of the UV−vis spectrum of hemoglobin

after the addition of 12, indicating that there is no interaction between

these two species.

PAMPA−BBB Assay. BBB penetration was evaluated by the

PAMPA−BBB assay, in which a porcine brain lipid was used as an

artiﬁcial membrane^28,39,40The ﬁve commercial drugs, phosphate buﬀer

saline (PBS, 10 mM), DMSO (for biology), and dodecane (analytical

standard,) were purchased from Sigma-Aldrich and Acros Organic.

The porcine brain lipid (PBL) was obtained from Avanti Polar Lipids

(100 mg, powder, catalogue no. 141101P). The donor plate was a 96-

well ﬁlter plate with hydrophobic polyvinylidene ﬂuoride (PVDF)

membrane (pore size 0.45 μm, nonsterile, catalogue no.

MAIPNTR10), and the acceptor plate was a 96-well transport receiver

plate (catalogue no. MATRNPS50), both from EDM Millipore Sigma.

The 96-well UV plate with a ﬂat bottom obtained from Greiner Bio-

One was used for UV measurements (catalogue no. 655801). Test

compounds were ﬁrst dissolved in DMSO to make a 10 mM stock

solution. Then 20 μL of the stock solution was diluted with 980 μL of

10 mM PBS buﬀer (pH = 7.5) to generate a ﬁnal concentration of 200

μM (2% DMSO). The acceptor plate was ﬁlled with 250 μL of 10 mM

PBS. The donor plate was ﬁrst coated with 4 μL of PBL (20 mg/mL in

dodecane). Then 250 μL of a test compound (200 μM) was

subsequently added to the donor plate. The donor plate was then

carefully placed on top of the acceptor plate to make a “sandwich”,

which was incubated at 25 °C for 17 h in a saturated humidity

atmosphere with an orbital agitation at 100 rpm. During this time,

compounds diﬀuse from the donor plate to the acceptor plate. After

incubation, 150 μL of test solution was taken from each well from both

sides (donor and acceptor) and transferred to the UV plate for

measurement. The concentration of a compound in each donor and

acceptor well was determined by using a standard curve, which was

built from its UV absorbance at λ_maxof various concentrations (1−200

μM). Each compound was measured in triplicate. The eﬀective

permeability (P_e) was calculated using the following equation:⁴¹

1-(3-(5-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)pyridin-3-yl)phenyl)-N-methylmethanamine (33b). Com-

pound 33b was prepared via Suzuki cross-coupling and reductive

amination. 32b was synthesized according to general procedure D

using Pd(PPh₃)₄(29.1 mg), K₂CO₃(143 mg), boronic acid 31b

(112.6 mg), and 30 (185 mg, 0.5 mmol). 33b was synthesized

according to general procedure E using N-methylamine hydrochloride

(103 mg), NaOAc (125 mg), 32b, and NaBH₄(116 mg). 33b was

isolated as a light-yellow oil (98.2 mg, 48%) after ﬂash column

chromatography (methanol:DCM 1:9). ¹H NMR (500 MHz, CDCl₃):

δ 8.65 (s, 1H), 8.36 (s, 1H), 7.66 (s, 1H), 7.51 (s, 1H), 7.40 (d, J = 5.0

Hz, 2H), 7.34 (t, J = 5.0 Hz, 1H), 6.89 (s, 1H), 6.85 (s, 1H), 5.86 (s,

2H), 3.82 (s, 2H), 3.13−3.12 (m, 4H), 2.47 (s, 3H), 2.34 (s, 3H), 2.08

(s, 6H). ¹³C NMR (125 MHz, CDCl₃): δ 160.1, 151.8, 149.7, 148.6,

146.1, 140.1, 138.0, 136.6, 136.1, 134.6, 129.2, 128.4, 128.1, 127.1,

126.0, 122.8, 120.4, 106.8, 55.6, 39.2, 35.7, 32.7, 21.0, 13.2.

1-(4-(5-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)pyridin-3-yl)phenyl)-N-methylmethanamine (33c). Com-

pound 33c was prepared via Suzuki cross-coupling and reductive

amination. 32c was synthesized according to general procedure D

using Pd(PPh₃)₄(29.3 mg), K₂CO₃(142 mg), boronic acid 31c (111.6

mg), and 30 (188 mg, 0.5 mmol). 33c was synthesized according to

general procedure E using N-methylamine hydrochloride (105 mg),

NaOAc (127 mg), 32c, and NaBH₄(119 mg). 33c was isolated as a

light-yellow oil (109.7 mg, 53%) after ﬂash column chromatography

1

(methanol:DCM 1:9). H NMR (500 MHz, CDCl₃): δ 8.64 (s, 1H),

8.37 (s, 1H), 7.62 (s, 1H), 7.46 (d, J = 8.0 Hz, 2H), 7.40 (d, J = 8.0

Hz, 2H), 6.88 (s, 1H), 6.85 (s, 1H), 5.87 (s, 2H), 3.79 (s, 2H), 3.13−

3.11 (m, 4H), 2.46 (s, 3H), 2.33 (s, 3H), 2.09 (s, 6H). ¹³C NMR (125

MHz, CDCl₃): δ 160.0, 151.8, 149.7, 148.5, 146.0, 139.9, 136.6, 136.5,

136.0, 134.4, 128.9, 128.4, 127.1, 122.8, 120.3, 106.8, 55.5, 39.2, 35.9,

32.7, 21.0, 13.2.

1-(3′-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)-5′-ﬂuoro-[1,1′-biphenyl]-3-yl)-N-methylmethanamine (35a).

Compound 35a was prepared via Suzuki cross-coupling and reductive

amination. 34a was synthesized according to general procedure D

using Pd(PPh₃)₄(28.8 mg), K₂CO₃(139 mg), boronic acid 31b

(113.6 mg), and 30 (186 mg, 0.5 mmol). 35a was synthesized

according to general procedure E using N-methylamine hydrochloride

(107 mg), NaOAc (131 mg), 34a, and NaBH₄(121 mg). 35a was

isolated as a light-yellow oil (141.9 mg, 66%) after ﬂash column

chromatography (methanol:DCM 1:9). ¹H NMR (500 MHz, CDCl₃):

δ 7.50 (s, 1H), 7.40 (d, J = 8.0 Hz, 1H), 7.37 (t, J = 7.5 Hz, 1H), 7.32

(d, J = 8.0 Hz, 1H), 7.19 (s, 1H), 7.11 (d, J = 7.5 Hz, 1H), 6.92 (s,

1H), 6.86 (s, 1H), 6.85 (d, J = 7.5 Hz, 1H), 5.88 (s, 2H), 3.81 (s, 2H),

3.12−3.11 (m, 4H), 2.48 (s, 3H), 2.36 (s, 3H), 2.11 (s, 6H); ¹³C NMR

(125 MHz, CDCl₃): δ 163.2 (d, J_C−F= 242.3 Hz), 160.5, 151.8, 149.6,

144.3 (d, J_C−F= 7.5 Hz), 143.1 (d, J_C−F= 8.8 Hz), 140.2 (d, J_C−F= 2.5

Hz), 140.1, 129.0, 128.5, 127.8, 127.0, 125.9, 123.1 (d, J_C−F= 1.3 Hz),

122.7, 120.3, 114.1 (d, J_C−F= 21.3 Hz), 111.7 (d, J_C−F= 22.5 Hz),

106.8, 55.8, 39.4, 35.8, 35.6, 21.0, 13.2.

1-(3′-(2-(6-(2,5-Dimethyl-1H-pyrrol-1-yl)-4-methylpyridin-2-yl)-

ethyl)-5′-ﬂuoro-[1,1′-biphenyl]-4-yl)-N-methylmethanamine (35b).

Compound 35b was prepared via Suzuki cross-coupling and reductive

amination. 34b was synthesized according to general procedure D

using Pd(PPh₃)₄(29.2 mg), K₂CO₃(143 mg), boronic acid 31c (115.3

mg), and 30 (186 mg, 0.5 mmol). 35b was synthesized according to

general procedure E using N-methylamine hydrochloride (109 mg),

NaOAc (135 mg), 34b, and NaBH₄(127 mg). 35b was isolated as a

light-yellow oil (127.5 mg, 60%) after ﬂash column chromatography

(methanol:DCM 1:9). ¹H NMR (500 MHz, CDCl₃): δ 7.48 (d, J = 8.5

Hz, 2H), 7.40 (d, J = 8.5 Hz, 2H), 7.14 (s, 1H), 7.07 (d, J = 9.5 Hz,

1H), 6.91 (s, 1H), 6.86 (s, 1H), 6.85 (d, J = 9.5 Hz, 1H), 5.88 (s, 2H),

3.82 (s, 2H), 3.11−3.10 (m, 4H), 2.48 (s, 3H), 2.35 (s, 3H), 2.10 (s,

6H). ¹³C NMR (125 MHz, CDCl₃): δ 163.2 (d, J_C−F= 242.5 Hz),

160.4, 151.8, 149.6, 144.3 (d, J_C−F= 7.5 Hz), 142.8 (d, J_C−F= 7.5 Hz),

139.2 (d, J_C−F= 2.5 Hz), 138.1, 129.1, 128.5, 127.2, 123.0 (d, J_C−F

=

M

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

original 90°. Therefore, a molecular replacement calculation with

PHASER-MR⁵⁰was needed to solve the structure. In the P2₁space

group, there are two heNOS dimers in the asymmetric unit.

Disordering in portions of inhibitors bound in the NOS active sites

was often observed, sometimes resulting in poor density quality.

However, partial structural features were usually still visible if the

contour level of the σA weighted 2m|F_o| − D|F_c| map was dropped to

0.5σ, which aﬀorded the building of reasonable models into the

disordered regions. Water molecules were added in PHENIX and

checked by Coot. The TLS⁵¹protocol was implemented in the ﬁnal

stage of reﬁnements with each subunit as one TLS group. The omit F_o

− F_cdensity maps were calculated by removing inhibitor coordinates

from the input PDB ﬁle before running one more round of TLS

reﬁnement in PHENIX (simulated annealing protocol with a 2000 K

initial temperature). The resulting map coeﬃcients DELFWT and

PHDELWT were used to generate maps. For some recent structures,

the Polder map facility in PHENIX was used to calculate the omit

density map for the bound inhibitors.⁵²The reﬁned structures were

validated in Coot before deposition in the Protein Data Bank.

⎡

⎤

V_A·V_D

A·(t − τ_ss) (V_A+ V_D)

V_A+ V_D

(1 − R)·V_D

C_A(t)

C_D(0)

2.303

P =

·

·lg 1 −

·

, where P_eis the

e

⎢

⎣

⎥

)

⎦

(

) (

eﬀective permeability (cm/s), V_Aand V_Dare the volume of the

acceptor and donor well (0.25 cm³), respectively, C_A(t) is the

concentration of the acceptor well at time t, C_D(0) and C_D(t) are the

concentration of the donor well at t₀and t, respectively, A is the ﬁlter

well area (0.21 cm²), and t is the incubation time (s). τ is the time to

ss

reach a steady state (usually very short compared to the incubation

time). R is the retention membrane factor and was calculated using the

⎡

C_D(t)

C_D(0)

V_A^C_A^(t)⎤

following equation: R = 1 −

−

·

. P_ewas reported as

⎢

⎣

⎥

⎦

V_DC_D(0)

an average of triplicate with a standard deviation.

Caco-2 Assay. The bidirectional Caco-2 assay was performed by

Sai Life Sciences, Pune, India. Caco-2 cells were ﬁrst grown in

Dulbecco’s Modiﬁed Eagles Medium until they attained 85−90%

conﬂuence. The cells were then trypsinized and seeded into 24-well

plates at a density of 0.6 × 10⁵cells/insert. The plates were kept in a

CO₂incubator at 37 °C for 21 days, and the medium was changed

every alternative day. TEER instrument was used to validate the

integrity of the monolayer. Wells with a TEER value above 230 Ω·cm²

were acceptable for the assay. Test compounds were dissolved in

DMSO to make a 5 mM stock solution, which was further diluted

1000-fold in Hank’s Balanced Salt Solution (HBSS) buﬀer to obtain a

ﬁnal concentration of 5 μM (0.1% DMSO). Studied compounds were

applied to either the apical (A → B direction) or the basal side (B → A

direction).Then 50 μL of test compounds in the receiver compartment

was taken at diﬀerent time points (0, 15, 30, 60, and 90 min) to

measure the permeability of compounds. The concentration of test

compounds at each time point was analyzed by LC-MS/MS. The

apparent permeability (P_app) was calculated using the following

equation: P_app= (dQ/dt)/C₀·A, where dQ/dt is the linear slope of

test compound concentration in the receiver chamber over time, C₀is

the initial concentration of the compounds in the donor well, and A is

the ﬁlter well area (0.7 cm²). The eﬄux ratio is deﬁned by the ratio of

the apparent permeability of B →A over that of A → B. An ER value

above 3 indicates that a compound is possibly a substrate of P-pg or

other active eﬄux transporters.

Inhibitor Complex Crystal Preparation. The sitting drop vapor

diﬀusion method was used to grow crystals at 4 °C for the heme

domains of rat nNOS (8 mg/mL containing 20 mM histidine), the

human nNOS K301R/R354A/G357D mutant (10 mg/mL), and

human eNOS (7 mg/mL). The crystal growth conditions were as

described previously.²²Fresh crystals were ﬁrst passed stepwise

through cryoprotectant solutions and then soaked with 5−10 mM

inhibitor for 3−4 h at 4 °C before being ﬂash cooled with liquid

nitrogen and stored until data collection. The presence of an acetate

ion near the heme active site in bovine eNOS caused interference in

the binding mode of some inhibitors.⁴²The high concentration of

magnesium acetate in the heNOS growth conditions may also

introduce an acetate near the active site that may inﬂuence the binding

mode of inhibitors. To avoid having this acetate in the structure, the

magnesium acetate in the cryoprotectant solution was replaced with

MgCl₂.

ASSOCIATED CONTENT

* Supporting Information

■

S

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acs.jmed-

chem.7b01356.

Crystallographic data collection and reﬁnement statistics

for rat nNOS, human nNOS, and human eNOS; crystal

structures of rnNOS hnNOS-7, rnNOS, hnNOS-13,

rnNOS-11, and rnNOS-19c; UV−vis absorption spec-

trum of hemoglobin without and in the presence of

compound 12; synthesis and analytical data for

compounds 20, 22a−d, 23a, 27, and 31a (PDF)

Molecular formula strings (CSV)

Accession Codes

PDB codes for X-ray structures described in this study are as

follow: rnNOS-7, 6AUQ; rnNOS-8, 6AUR; rnNOS-10, 6AUS;

rnNOS-12, 6AUT; rnNOS-13, 6AUU; rnNOS-16, 6AUV;

rnNOS-17, 6AUW; rnNOS-18, 6AUX; hnNOS-7, 6AUY;

hnNOS-8, 6AUZ; hnNOS-10, 6AU0; hnNOS-12, 6AU1;

hnNOS-13, 6AU2; hnNOS-16, 6AU3; hnNOS-17, 6AU4;

hnNOS-18, 6AU5; heNOS-8, 6AU6; heNOS-12, 6AU7.

Authors will release the atomic coordinates and experimental

data upon article publication.

AUTHOR INFORMATION

Corresponding Authors

*For R.B.S.: phone, +1 847 491 5653; fax: +1 847 491 7713; E-

mail, Agman@chem.northwestern.edu.

■

*For T.L.P.: phone, +1 949 824 7020; E-mail, poulos@uci.edu.

X-ray Diﬀraction Data Collection, Data Processing, and

Structural Reﬁnement. The cryogenic (100 K) X-ray diﬀraction

data were collected remotely at the Stanford Synchrotron Radiation

Lightsource (SSRL) or Advanced Light Source (ALS) through the

data collection control software Blu-Ice⁴³and a crystal-mounting

robot. When a CCD detector was used, 100−125° of data were

typically collected with 0.5° per frame. If a Pilatus pixel array detector

was used, 140−160° of ﬁne-sliced data were collected with a 0.2° per

frame. Raw CCD data frames were indexed, integrated, and scaled

using iMOSFLM,⁴⁴but the pixel array data were processed with

XDS⁴⁵and scaled with Aimless.⁴⁶The binding of inhibitors was

detected by initial diﬀerence Fourier maps calculated with REFMAC.⁴⁷

The inhibitor molecules were then modeled in Coot⁴⁸and reﬁned

using REFMAC or PHENIX.⁴⁹The crystal packing of the MgCl₂

soaked heNOS crystals was changed slightly, resulting in a symmetry

change from the orthorhombic P2₁2₁2₁reported previously³⁶to

monoclinic P2₁, with a β angle only 0.6−0.7° oﬀ compared to the

ORCID

Thomas L. Poulos: 0000-0002-5648-3510

Richard B. Silverman: 0000-0001-9034-1084

Author Contributions

^§H.T.D and H.-Y.W. contributed equally.

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

■

We are grateful for the generous support from the National

Institutes of Health (R01 GM049725, to R.B.S., GM057353 to

T.L.P.). H.L. thanks Carla Plaza for her assistance in NOS

protein expression and puriﬁcation; the puriﬁed samples were

used in both crystallography and enzyme assays. We also wish

N

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(15) Silverman, R. B. Design of selective neuronal nitric oxide

synthase inhibitors for the prevention and treatment of neuro-

degenerative diseases. Acc. Chem. Res. 2009, 42, 439−451.

(16) Di, L.; Rong, H.; Feng, B. Demystifying brain penetration in

central nervous system drug discovery. J. Med. Chem. 2013, 56, 2−12.

(17) Mukherjee, P.; Cinelli, M. A.; Kang, S.; Silverman, R. B.

Development of nitric oxide synthase inhibitors for neurodegeneration

and neuropathic pain. Chem. Soc. Rev. 2014, 43, 6814−6838.

(18) Labby, K. J.; Xue, F.; Kraus, J. M.; Ji, H.; Mataka, J.; Li, H.;

to thank the SSRL and ALS beamline staﬀ for their support

during remote X-ray diﬀraction data collection. This work made

use of the IMSERC at Northwestern University, which has

received support from the Soft and Hybrid Nanotechnology

Experimental (SHyNE) Resource (NSF NNCI-1542205), the

State of Illinois, and and the International Institute for

Nanotechnology (IIN).

́

Martasek, P.; Roman, L. J.; Poulos, T. L.; Silverman, R. B.

ABBREVIATIONS USED

■

Intramolecular hydrogen bonding: a potential strategy for more

bioavailable inhibitors of neuronal nitric oxide synthase. Bioorg. Med.

Chem. 2012, 20, 2435−2443.

NO, nitric oxide; nNOS, neuronal nitric oxide synthase; iNOS,

inducible nitric oxide synthase; eNOS, endothelial nitric oxide

synthase; hnNOS, human neuronal nitric oxide synthase;

heNOS, human endothelial nitric oxide synthase; ^L-Arg, L-

arginine; FAD, ﬂavin adenine dinucleotide; FMN, ﬂavin

mononucleotide; NADPH, reduced nicotinamide adenine

dinucleotide phosphate; H₄B, (6R)-5,6,7,8-tetrahydrobiopterin;

HEPES, 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid;

DNA, eoxyribonucleic acid; WT, wild type; PAMPA, parallel

artiﬁcial membrane permeability assay; BBB, blood−brain

barrier; CNS, central nervous system; P-gp, P-glycoprotein;

ATP, adenosine triphosphate; ER, eﬄux ratio; P_e, eﬀective

permeability; P_app, apparent permeability.

(19) Silverman, R. B.; Lawton, G. R.; Ranaivo, H. R.; Chico, L. K.;

Seo, J.; Watterson, D. M. Effect of potential amine prodrugs of

selective neuronal nitric oxide synthase inhibitors on blood−brain

barrier penetration. Bioorg. Med. Chem. 2009, 17, 7593−7605.

́

(20) Xue, F.; Li, H.; Delker, S. L.; Fang, J.; Martasek, P.; Roman, L. J.;

Poulos, T. L.; Silverman, R. B. Potent, highly selective, and orally

bioavailable gem-difluorinated monocationic inhibitors of neuronal

nitric oxide synthase. J. Am. Chem. Soc. 2010, 132, 14229−14238.

́

(21) Mukherjee, P.; Li, H.; Sevrioukova, I.; Chreifi, G.; Martasek, P.;

Roman, L. J.; Poulos, T. L.; Silverman, R. B. Novel 2,4-disubstituted

pyrimidines as potent, selective, and cell-permeable inhibitors of

neuronal nitric oxide synthase. J. Med. Chem. 2015, 58, 1067−1088.

(22) Cinelli, M. A.; Li, H.; Chreifi, G.; Poulos, T. L.; Silverman, R. B.

Nitrile in the hole: discovery of a small auxiliary pocket in neuronal

nitric oxide synthase leading to the development of potent and

selective 2-aminoquinoline inhibitors. J. Med. Chem. 2017, 60, 3958−

3978.

REFERENCES

■

(1) Garthwaite, J.; Boulton, C. L. Nitric oxide signaling in the central

nervous system. Annu. Rev. Physiol. 1995, 57, 683−706.

(2) Bogdan, C. Nitric oxide and the immune response. Nat. Immunol.

2001, 2, 907−916.

(3) Marin, J.; Rodriguezmartinez, M. A. Nitric oxide, oxygen-derived

free-radicals and vascular endothelium. J. Auton. Pharmacol. 1995, 15,

279−307.

(4) Alderton, W. K.; Cooper, C. E.; Knowles, R. G. Nitric oxide

synthases: structure, function and inhibition. Biochem. J. 2001, 357,

593−615.

(5) Knowles, R. G.; Moncada, S. Nitric oxide synthases in mammals.

Biochem. J. 1994, 298, 249−258.

(6) Maccallini, C.; Amoroso, R. Targeting neuronal nitric oxide

synthase as a valuable strategy for the therapy of neurological

disorders. Neural Regener. Res. 2016, 11, 1731−1734.

(7) Heinrich, U. R.; Helling, K. Nitric oxide - a versatile key player in

cochlear function and hearing disorders. Nitric Oxide 2012, 27, 106−

116.

́

(23) Wang, H.-Y.; Qin, Y.; Li, H.; Roman, L. J.; Martasek, P.; Poulos,

T. L.; Silverman, R. B. Potent and selective human neuronal nitric

oxide synthase inhibition by optimization of the 2-aminopyridine-

based scaffold with a pyridine linker. J. Med. Chem. 2016, 59, 4913−

4925.

́

(24) Kang, S.; Li, H.; Tang, W.; Martasek, P.; Roman, L. J.; Poulos,

T. L.; Silverman, R. B. 2-aminopyridines with a truncated side chain to

improve human neuronal nitric oxide synthase inhibitory potency and

selectivity. J. Med. Chem. 2015, 58, 5548−5560.

(25) Rankovic, Z. CNS drug design: balancing physicochemical

properties for optimal brain exposure. J. Med. Chem. 2015, 58, 2584−

2608.

(26) Hevel, J. M.; Marletta, M. A. [25] Nitric-oxide synthase assays.

Methods Enzymol. 1994, 233, 250−258.

(27) Yung-Chi, C.; Prusoff, W. H. Relationship between the

inhibition constant (Ki) and the concentration of inhibitor which

causes 50% inhibition (I50) of an enzymatic reaction. Biochem.

Pharmacol. 1973, 22, 3099−3108.

(8) Dusting, G. J. Nitric oxide in cardiovascular disorders. J. Vasc. Res.

1995, 32, 143−161.

̀

(9) Mollace, V.; Nistico, G. Release of nitric oxide from astroglial

(28) Di, L.; Kerns, E. H.; Fan, K.; McConnell, O. J.; Carter, G. T.

High throughput artificial membrane permeability assay for blood−

brain barrier. Eur. J. Med. Chem. 2003, 38, 223−232.

(29) Di, L.; Kerns, E. H.; Bezar, I. F.; Petusky, S. L.; Huang, Y.

Comparison of blood−brain barrier permeability assays: in situ brain

perfusion, MDR1-MDCKII and PAMPA-BBB. J. Pharm. Sci. 2009, 98,

1980−1991.

cells: a key mechanism in neuroimmune disorders. Adv. Neuroimmunol.

1995, 5, 421−430.

(10) de la Torre, J. C.; Stefano, G. B. Evidence that Alzheimer’s

disease is a microvascular disorder: the role of constitutive nitric oxide.

Brain Res. Rev. 2000, 34, 119−136.

́

(11) Drechsel, D. A.; Estevez, A. G.; Barbeito, L.; Beckman, J. S.

́

Nitric oxide-mediated oxidative damage and the progressive demise of

motor neurons in ALS. Neurotoxic. Res. 2012, 22, 251−264.

(12) Uehara, T.; Nakamura, T.; Yao, D.; Shi, Z.-Q.; Gu, Z.; Ma, Y.;

Masliah, E.; Nomura, Y.; Lipton, S. A. S-Nitrosylated protein-

disulphide isomerase links protein misfolding to neurodegeneration.

Nature 2006, 441, 513−517.

(30) Konczol, A.; Muller, J.; Foldes, E.; Beni, Z.; Vegh, K.; Kery, A.;

́

̈

Balogh, G. T. Applicability of a blood−brain barrier specific artificial

membrane permeability assay at the early stage of natural product-

based CNS drug discovery. J. Nat. Prod. 2013, 76, 655−663.

(31) Patil, A. G.; D’Souza, R.; Dixit, N.; Damre, A. Validation of

quinidine as a probe substrate for the in vitro P-gp inhibition assay in

Caco-2 cell monolayer. Eur. J. Drug Metab. Pharmacokinet. 2011, 36,

115.

́

(13) Torreilles, F.; Salman-Tabcheh, S. D.; Guerin, M.-C.; Torreilles,

J. Neurodegenerative disorders: the role of peroxynitrite. Brain Res.

Rev. 1999, 30, 153−163.

(14) Bolanos, J. P.; Almeida, A.; Stewart, V.; Peuchen, S.; Land, J. M.;

Clark, J. B.; Heales, S. J. R. Nitric oxide-mediated mitochondrial

damage in the brain: mechanisms and implications for neuro-

degenerative diseases. J. Neurochem. 1997, 68, 2227−2240.

(32) Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. Moving

beyond rules: the development of a central nervous system

multiparameter optimization (CNS MPO) approach to enable

alignment of druglike properties. ACS Chem. Neurosci. 2010, 1,

435−449.

O

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

(33) Di, L.; Kerns, E. H. Drug-Like Properties: Concepts, Structure

Design, and Methods from ADME to Toxiciy Optimization, 2nd Ed.,

Elsevier: Amsterdam, 2016; p 329.

(34) Gerber, N. C.; De Montellano, P. R. O. Neuronal nitric oxide

synthase: expression in Escherichia coli, irreversible inhibition by

phenyldiazene, and active site topology. J. Biol. Chem. 1995, 270,

17791−17796.

(35) Hevel, J. M.; White, K. A.; Marletta, M. A. Purification of the

inducible murine macrophage nitric oxide synthase. Identification as a

flavoprotein. J. Biol. Chem. 1991, 266, 22789−22791.

(36) Li, H.; Jamal, J.; Plaza, C.; Pineda, S. H.; Chreifi, G.; Jing, Q.;

Cinelli, M. A.; Silverman, R. B.; Poulos, T. L. Structures of human

constitutive nitric oxide synthases. Acta Crystallogr., Sect. D: Biol.

Crystallogr. 2014, 70, 2667−2674.

(51) Winn, M. D.; Isupov, M. N.; Murshudov, G. N. Use of TLS

parameters to model anisotropic displacements in macromolecular

refinement. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2001, 57, 122−

133.

(52) Liebschner, D.; Afonine, P. V.; Moriarty, N. W.; Poon, B. K.;

Sobolev, O. V.; Terwilliger, T. C.; Adams, P. D. Polder maps:

improving OMIT maps by excluding bulk solvent. Acta Crystallogr.,

Sect. D: Struct. Biol. 2017, 73, 148−157.

(37) Roman, L. J.; Sheta, E. A.; Martasek, P.; Gross, S. S.; Liu, Q.;

Masters, B. S. High-level expression of functional rat neuronal nitric

oxide synthase in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 1995,

92, 8428−8432.

(38) Leber, A.; Hemmens, B.; Klosch, B.; Goessler, W.; Raber, G.;

̈

Mayer, B.; Schmidt, K. Characterization of recombinant human

endothelial nitric-oxide synthase purified from the yeast Pichia

pastoris. J. Biol. Chem. 1999, 274, 37658−37664.

̌ ̌ ̌

́ ̌ ́ ́ ́

L.; Perez, D. I.; Stepankova, S.; Chlebek, J.; Safratova,

́

A.; Opletal, L. In vitro inhibitory effects of 8-O-

(39) Cahlíkova,

́

M.; Host’alkova,

̌

́

demethylmaritidine and undulatine on acetylcholinesterase and their

predicted penetration across the blood−brain barrier. J. Nat. Prod.

2015, 78, 1189−1192.

́ ́

(40) Teixido, M.; Zurita, E.; Malakoutikhah, M.; Tarrago, T.; Giralt,

E. Diketopiperazines as a tool for the study of transport across the

blood−brain barrier (BBB) and their potential use as BBB-shuttles. J.

Am. Chem. Soc. 2007, 129, 11802−11813.

́

̋ ́

(41) Muller, J.; Esso, K.; Dargo, G.; Konczol, A.; Balogh, G. T.

̈ ̈

̈

Tuning the predictive capacity of the PAMPA-BBB model. Eur. J.

Pharm. Sci. 2015, 79, 53−60.

(42) Cinelli, M. A.; Li, H.; Pensa, A. V.; Kang, S.; Roman, L. J.;

́

Martasek, P.; Poulos, T. L.; Silverman, R. B. Phenyl ether- and aniline-

containing 2-aminoquinolines as potent and selective inhibitors of

neuronal nitric oxide synthase. J. Med. Chem. 2015, 58, 8694−8712.

(43) McPhillips, T. M.; McPhillips, S. E.; Chiu, H.-J.; Cohen, A. E.;

Deacon, A. M.; Ellis, P. J.; Garman, E.; Gonzalez, A.; Sauter, N. K.;

Phizackerley, R. P.; Soltis, S. M.; Kuhn, P. Blu-ice and the distributed

control system: software for data acquisition and instrument control at

macromolecular crystallography beamlines. J. Synchrotron Radiat. 2002,

9, 401−406.

(44) Battye, T. G. G.; Kontogiannis, L.; Johnson, O.; Powell, H. R.;

Leslie, A. G. W. iMOSFLM: a new graphical interface for diffraction-

image processing with MOSFLM. Acta Crystallogr., Sect. D: Biol.

Crystallogr. 2011, 67, 271−281.

(45) Kabsch, W. XDS. Acta Crystallogr., Sect. D: Biol. Crystallogr.

2010, 66, 125−132.

(46) Evans, P. Scaling and assessment of data quality. Acta

Crystallogr., Sect. D: Biol. Crystallogr. 2006, 62, 72−82.

(47) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of

macromolecular structures by the maximum-likelihood method. Acta

Crystallogr., Sect. D: Biol. Crystallogr. 1997, 53, 240−255.

(48) Emsley, P.; Cowtan, K. Coot: model-building tools for

molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,

60, 2126−2132.

(49) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis,

I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-

Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.

J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H.

PHENIX: a comprehensive Python-based system for macromolecular

structure solution. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2010, 66,

213−221.

(50) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.

D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl.

Crystallogr. 2007, 40, 658−674.

P

DOI: 10.1021/acs.jmedchem.7b01356

J. Med. Chem. XXXX, XXX, XXX−XXX

Article Doi

Improvement of Cell Permeability of Human Neuronal Nitric Oxide Synthase Inhibitors Using Potent and Selective 2-Aminopyridine-Based Scaffolds with a Fluorobenzene Linker

DOI: 10.1021/acs.jmedchem.7b01356

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.7b01356

Full text of DOI:10.1021/acs.jmedchem.7b01356

Products guided by the article

R&D Labs maybe for 817618-57-8

Relevant to this article

Hot Product